^\ 
 
 .1 CV M 
 
 J^.^ 
 
 rm% 
 
 pt/j 
 
 II 
 
 .¥ 
 
 ifef *■ 
 
 .^ ^ m 
 
 ■<^yi 
 
 ^% 
 
®l|p i. Hi. BtU Slibrara 
 
 Nortli (EarflUna ^tatt Imoeraita 
 
 QH406 
 
 ia3 
 
THIS BOOK IS DUE ON THE DATE 
 INDICATED BELOW AND IS SUB- 
 JECT TO AN OVERDUE FINE AS 
 POSTED AT THE CIRCULATION 
 DESK. 
 
 ib.^ 
 
 ti/i f^"-'*'-' •*''"*^"" 
 
North Carolina State Library 
 Raleigh 
 
 MUTATIONS, VARIATIONS, AND RELATION- 
 SHIPS OF THE OENOTHERAS 
 
 BY 
 
 D. T. MACDOUGAL, A. M. VAIL. AND G. H. SHULL 
 
 WASHINGTON, D. C: 
 Published by the Carnegie Institution of Washington 
 
 1907 
 
North Carolina State Library 
 
 Gift of 
 
North CaroliDMRate ua^aty 
 
 MUTATIONS, VARIATIONS, AND RELATION- 
 SHIPS OF THE OENOTHERAS 
 
 BY 
 
 D. T. MACDOUGAL, A. M. VAIL, AND G. H. SHULL 
 
 WASHINGTON, D. C. : 
 
 Published by tlie Carnegie Institution ot Washington 
 
 1907 
 
/ 
 
 / 
 
 Carnegie Institution of Washington Publication No. 8i. 
 Papers of the Station for Experimental Evolution, No. q. 
 
PLATE 1 
 
 nx^' 
 
 K,-^ 
 
 J ? 
 
 
 
 c _ - 
 
 •/. S '- 
 
 •r 3 -5 Sv 
 
 1 :!■ ^- 'C 
 
 /, :- ■/ 5^ 
 
MUTATIONS, VARIATIONS. AND RliLATIONSHlPS 
 
 OP THE OHNOTHHRAS. 
 
 By D. T. MacDougau, A. M. Vail, and G. H. Shull. 
 
 SCOPE OF INVESTIGATION. 
 
 The Oenotheras have furnished so much evidence of importance in connec- 
 tion with saltatory action in heredity that it has been deemed important to 
 continue the cultural investigation of the group begun in 1902. 
 
 Sseds representing the spscies in cultivation in the principal botanical 
 gardens of the world have been procured, and these, in addition to a large num- 
 ber of forms native to eastern North America, have been grown in guarded 
 cultures. 
 
 Attention has bsen paid to the occurrence of mutants in Oenothera lamarck- 
 ian2 with a view to testing: the coefficient of mutabilitv and the influence of 
 environmental conditions on mutation. Extensive sowings have been made 
 for the purpose of finding derivatives hitherto undetected, with a coefficient of 
 mutability so small as to have escaped observation. Descriptions of known 
 mutants have been made independently for the purpose of comparison with 
 supposedly identical forms in Amsterdam and of facilitating observations 
 of all kinds upon the subject. 
 
 Many important relations between mutants and their parents may be most 
 advantageously considered by statistical methods, and the studies begun by 
 one of the authors in 1904 have been continued and extended to include addi- 
 tional mutants. The height and branching of the stems and the width and 
 length of the leaves have been again taken into account, but owing to the 
 great susceptibility of these organs to variation in direct response to environ- 
 ment, measurements have also been made upon the buds. The lesser varia- 
 tion of the latter in correlation with vegetative characters makes them much 
 more satisfactory for the study of hereditary relations, and it is clear that 
 their statistical study in connection with pedigree-cultures will demonstrate 
 in several generations the permanence or evanescence of the mutant types 
 and give decisive answers to such questions as the relation between fluctua- 
 tion and mutation and the "fixing" of variations through self-fertihzations 
 or their disappearance through crossings. 
 
2 MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 
 
 Guarded psdigree-cultures have been made from pure seeds of native species 
 of evening- primroses which had not previously been brought under close 
 observation, with a view to procuring additional evidence on mutabiHty in 
 this group. 
 
 The preliminary examination made it apparent that the group comprises a 
 swarm of clearlv separable species, many of which had not been recognized 
 by the taxonomists and which gave basis for the current opinion as to the 
 wide variability of the common evening-primrose (D. biennis). 
 
 The studv of genetic relationships and of the phenomena of hybridization 
 in general has necessitated the organization of extensive cultures, which wall 
 need continuance for two or more seasons before definite results may be 
 announced, and their discussion is reserved for a future paper. 
 
 A few cases of fixed hybrids have been encountered, however, in which the 
 crossing of two species resulted in the production of a progeny consisting of 
 one fixed tvpe constant in successive generations. Forms arising in this way 
 are, in reality, new species, and doubtless many such have arisen naturally 
 and now constitute a part of the native flora. 
 
 Striking cases of vegetative mutation, or bud-sports, having arisen, atten- 
 tion has been devoted to a study of the inheritance of the saltatory groups of 
 characters. 
 
 A systematic attempt to localize mutatory changes in the life-cycle of the 
 sporoph}^e has been made, and has met with marked success in a preliminary 
 wav, since it has been possible to induce new mutants by the use of solutions 
 of strong osmotic activity and by highly dilute preparations of mineral salts, 
 some of which are poisonous to plants in high concentrations and stimulative 
 in low concentrations. Finally, an attempt has been made to bring the facts 
 disclosed in the descriptive part of the paper into correlation with prevailing 
 theories as to phylogeny and evolutionary procedure. 
 
 The principal results were obtained from cultural investigations in the 
 experimental grounds and greenhouses of the New York Botanical Garden 
 and of the Station for Experimental Evolution at Cold Spring Harbor, Long 
 Island, and were also extended in the later stages to the Desert Botanical 
 Laboratory of the Carnegie Institution of Washington at Tucson, Arizona, 
 thus securing the advantages of a wide range of climate and soils. 
 
 A note regarding the various mutants was presented at the weekly botanical 
 convention at the New York Botanical Garden, November i, 1905, and a 
 second one, dealing especially with sports or vegetative mutations, before the 
 Torrey Botanical Club, on November 4, 1905. A lecture dealing with his- 
 torical aspects of the phases of the subject under investigation was presented 
 to the Barnard Botanical Club, December 18, 1905, in which the first announce- 
 ment was made of the induction of mutants by chemically and osmotically 
 active stimuli. A full description of the technique of pedigree-cultures and 
 
MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 3 
 
 of the methods emploved in the stimulation of ovaries was given in a lecture 
 at the Woods Hole Marine Biological Laboratory, July 20, 1906, in conjunction 
 with which derivatives thus obtained of the second generation in Raimannia 
 and of the first of 0. biennis were exhibited. The results of Dr. Shull's 
 statistical inquiries were presented before the Botanical Society of America at 
 New Orleans, January 3, 1906, and various notes have been presented before 
 other societies. 
 
 PEDIGREE-CULTURES OF OENOTHERA LAMARCKIANA. 
 
 Arrangements were made by which seeds were obtained from four separate 
 individuals, purely fertilized with their own pollen. This was done by apply- 
 ing the pollen to the pistil of the same flower or of another flower of the same 
 plant. All sowings were made in earthenware pans, 30 by 30 cm., filled with 
 soil sterilized in an autoclave as in previous cultures, and when sufficiently 
 developed the seedlings were removed, to be transplanted or discarded, as 
 the conditions of the experiments demanded. The cultures are given under 
 the key-numbers by which they were designated in the journal of the experi- 
 mental garden. 
 
 A. I. 0. — -A packet of seeds which was given the foregoing designation, 
 which had bsen harvested in Amsterdam in 1901, and was obtained directly 
 from Professor Ds Vries. The sowing was made in August, 1904, and the 
 small rosettes were inspected by Professor De Vries late in September, 1904, 
 and he kindly assisted in the identification of a few of the mutants included. 
 Some confusion in the record makes it impossible to give the exact census of 
 the culture, but it comprised between 500 and 600 seedlings, among which 
 26 mutant derivatives were identifiable, and, so far as possible, two represent- 
 atives of each type were transplanted to the experimental garden in May, 
 1905, coming into bloom about 60 days later. 
 
 The authors have not had the opportunity of inspecting living specimens 
 of all of the mutants which have appeared in Amsterdam and which have 
 been described by Professor De Vries, and of the 9 forms seen but 5 have 
 been conclusively recognized. Of these, O. oblonga was represented by 12 
 individuals, constituting 46 per cent of the total number of mutants. The 
 average frequency of this type among the mutative progeny is i per cent, or 
 20 per cent of the mutants, although in one series De Vries found 176 ohlongas 
 among a total of 334 mutants derived from a culture consisting of over 14,000 
 individuals. In this instance oblonga constituted nearly 53 per cent of the 
 mutants or 1.25 per cent of the entire progeny. (In De Vries, 1905, p. 545, 
 the proportion of oblonga in cultures is given as 10 per cent, which is a mis- 
 print for I per cent.) 
 
 In the New York culture under discussion, lata was represented by one 
 individual, nanella by one, albida by two, and gigas by one. Of these, the 
 
4 MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE) OENOTHERAS. 
 
 appearance of gigas is the most notable, since it has occurred but few times 
 in the last twenty years in pedigree-cultures of 0. lamarckiana. 
 
 The remaining four forms among the mutants were represented by one, 
 one, two, and five individuals. Of these, one type with orbicular leaves 
 presented a most striking departure from the parental form, but failed to 
 perfect flowers, so that but little information concerning it was gained. The 
 form in question was reprseented by one individual only, and did not appear 
 in anv other culture 
 
 It is to be noted that the seeds used in this culture had been stored for three 
 years. Doubtless the parental type and the derivatives are characterized 
 by varying and different powers of endurance, so that the coefficient of fre- 
 quency of the several forms in question might be expected to be slightly 
 dift'erent from that found in freshly harvested seeds. 
 
 C.I. 2. — A lot of purely fertilized seeds derived from one individual, har- 
 vested in the New York Botanical Garden in 1903, was sown in sterilized soil 
 in the greenhouse in August, 1904. In September the plantlets w^ere so far 
 advanced as to admit of recognition of a number of forms, and a number of 
 duplicates of the parental type were discarded. This process was repeated 
 at intervals throughout the winter, and the remaining rosettes were trans- 
 planted to the experimental garden in May, 1905. vSeven derivatives were 
 seen, including 33 individuals. Among these, were 2 of scintillans, 3 of albida, 
 18 of oblonga, and 2, 2, 3, and 3 of four other unknown types. Some of these 
 unknown types also occurred in other cultures, and 3 of the known and 
 unknown were not duplicated in other progenies. 
 
 The total number of the seedlings included was probablv not much above 
 500, and if this were conclusively confirmed it would furnish an illustration of 
 a progeny in which the mutants constituted about 6 per cent of the entire 
 number. Unfortunately some confusion was found in the record. Before 
 this was discovered, the senior author announced that he had succeeded in 
 modifying the coefficient of mutability, but, as is to be seen from the above, 
 the evidence is not conclusive. The large proportion of oblonga is noticeable, 
 and this form constitutes 54 per cent of the total number of the mutants. 
 
 D. I. 7. — A lot of purely fertilized seeds taken from one individual, grown 
 in the New York Botanical Garden in 1904, was sown in the garden in August, 
 1904, furnishing a total of 499 plantlets. Development proceeded so rapidly 
 that it was possible to select and discard 244 duplicates of the parental type 
 on December 4, 1904. The remaining ones were transferred to small pots. 
 Forty-three more duplicates were discarded on Februarv 9, 1Q05, 34 on Feb- 
 ruary 18, 40 on March 3, 6 on INIarch 21, 2 on March 24, 3 on March 27, 7 on 
 March 30, and i on April 17. On May 15, 19 mutants and 2 of the parental 
 type were transplanted to the experimental garden. The detail given above 
 illustrates the general procedure in all such cultures. 
 
MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 5 
 
 As a result of the continued inspection of the cuhure it was seen to include 
 3 of nanella, 2 of lata, 6 of albida, and 3 of oblonga, while the remainder could 
 not be identified. One was purely pollinated to test its relationship with 
 gigas. The coefficient of mutability in this culture was seen to be barely 3 
 per cent, and the proportion of oblonga was much lower than in any other 
 culture. 
 
 D. I. 9. — A lot of purely fertilized seeds designated as above, harvested in 
 the New York Botanical Garden in August, 1904, was sown in two earthen- 
 ware pans filled with sterilized soil, early in November, 1904; 604 plantlets 
 were produced ; 390 duplicates of the parental type were discarded on Decem- 
 ber 4th, 1904, and 66 on December 15th; 77 were transplanted for further 
 inspection on December 21st, and the remaining 71 on January 15th, 1905; of 
 these, 59 were discarded on February 9th, 33 on February i8th, 15 on March 3d, 
 3 on March 21st, 5 on jMarch 24th, and 4 on March 30th; 19 mutant individuals 
 were transplanted to the experimental garden on May i6th, 1905, of which 
 I was nanella, 2 were lata, 2 scintillans, i albida, and 2 oblonga, and the 
 remainder were of types not identified by the authors. 
 
 The coefficient of mutability in this instance is seen to be about 3 per cent, 
 and that of oblonga very small in comparison with the general frequency of 
 this derivative. 
 
 IDENTITY AND DISTRIBUTION OF OENOTHERA LAMARCKIANA. 
 
 Since the cultures of the evening-primroses was begun, a few years since, 
 no opportunity has been neglected to attempt to trace 0. lamarckiana to its 
 original habitat and to establish its relationship to other species of the genus. 
 Among numerous bibliographical discussions of interest in this connection, 
 one by Miller (1760) is of great interest. He says regarding the "Tree Prim- 
 rose with oval spear-shaped indented Leaves, and Flowers proceeding from 
 the wings of the Teaves on the Upper Part of the Stalk:" 
 
 This plant is also a Native of North America; but was the first species of the Genus which 
 was brought to Europe, so it is more commonly seen in the Gardens than any of the other 
 species. In some parts of Europe, this is spread about from the Gardens in such plenty, 
 that it might be supposed a native there. In a small wood near Haarlem in Holland, this 
 plant covered the ground insomuch that n^any skilful persons supposed it was a native of 
 that place. But it may be easily accounted for, because the gardeners who live near that 
 place are chiefly .^orists, and they annually change the earth of the beds in their gardens; 
 so by carrying out of their old earth from their beds, in which many of the seeds were 
 scattered, the plants came up there; and those being suffered to scatter their seeds, had 
 filled the whole wood with the plants. 
 
 This differs from tlie first sort (described and figured as 0. biennis) in having broader 
 leaves; the stalk grows taller, and the flowers are much larger. Both these sorts will thrive 
 in the Smoak of London better than most plants. 
 
 The appended descriptions and the plate (No. 189, dated 1757) very fit- 
 tingly characterize O. biennis and 0. lamarckiana, and as the descriptions were 
 
6 MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 
 
 made before 0. grandi flora, the only known species which might be confused 
 with 0. lamarckiana, was discovered, the inference may be drawn by exclu- 
 sion that nothing but O. lamarckiana was referred to, and this inference is 
 strengthened by the fact that 0. lamarckiana was recognized and cultivated 
 in the Paris gardens a few years later. 
 
 Of the cultures that have been carried out from material obtained from 
 various sources, the following are selected as of unusual interest : 
 
 Two sheets of dried material of 0. lamarckiana, collected by Mr. E. P. Bick- 
 nell at Nantucket City, were received in September, 1904. These plants had 
 been found in the previous month, growing near a cottage, and had spread to 
 an adjoining waste lot. A visit was made to the place in August, 1905, by 
 the senior author, but no trace of the plant could be seen, although it was 
 found again in 1906 by Mr. Bicknell. This journey was inspired by the behav- 
 ior of the seeds taken from the herbarium specimen and sown in sterilized soil 
 in November, 1904. No particular attention v/as paid to the lot, but on 
 January 27, 1904, 24 plantlets representing the widest diversity observable 
 were transplanted to small pots in accordance with the usual practice. Six of 
 these corresponded quite exactly to the mutant 0. albida. Four of these 
 died before May, 1905, at which time the remainder were transplanted to the 
 experimental grounds. All of the other individuals developed in accordance 
 with qualities of 0. lamarckiana, with a maximum amount of color in the buds, 
 and also a maximum number of basal branches of some length. The pale 
 mutants, however, did not proceed beyond the rosette stage in 1905. 
 
 The owner of the grounds in which the original plants were growing in 
 Nantucket could not give any information as to the origin of the culture, 
 except to say that it had been started from seeds a great many years before. 
 
 A lot of seeds taken from a trade packet sold by Vilmorin & Company, of 
 Paris, were sown in sterilized soil on Januar}^ 3, 1905. The record shows that 
 192 duplicates of the parental type had been discarded when the 1 1 remaining 
 ones were transplanted to the experimental garden on May 16. Of these but 
 3 were mutants, including i lata, i nanella, and i albida. The mutants were 
 thus seen to constitute but 1.5 per cent of the total culture. 
 
 Far the most interesting material and records were obtained from England, 
 and it was found that 0. lamarckiana has been growing in some profusion 
 in certain localities in that country for several years. 
 
 In 1905 Mr. H. Stuart Thompson (Thompson, 1905) called attention to 
 Oenothera biennis L. and 0. odorata Jacq., growing on sandhills in Lancashire 
 and Somersetshire, and with the view of comparing the native North American 
 0. biennis with the plant so called growing in England, Mr. C. Theodore Green 
 was applied to for seeds of the plant growing in Liverpool district. To him, 
 therefore, we are indebted for seeds, a photograph of the Lancashire 0. biennis, 
 and the note on the following page (Green, 1902). 
 
MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 7 
 
 Oenothera "biennis" Linn^us. 
 
 Form as found on both sides of the River Mersey, from the estuary of the River Dee to 
 Southport, 1 8 miles north of Liverpool. It is especially luxuriant about Formby Sandhills, 
 9 miles north of Liverpool, and about Bidston railway junction, 3 miles north of Birken- 
 head. It flourishes in sandy waste ground, chiefly among the hills of blown sand along the 
 coast, and extends a few miles inland in scattered groups of plants. It also is found fre- 
 quently in cultivated ground in cottage and other gardens from the Dee to Southport. I 
 do not know the date of its first appearance in the Liverpool district, but it was mentioned 
 in Dickinson's Flora of Liverpool publication, 1851 and 1855. 
 
 F. W. Webb saw it near Leasowe (i860), 5 miles southwest of Liverpool, and T. Sansom 
 at New Brighton (Dick, Flora). 
 
 I have known it in various parts of this peninsula of W'irral (between Mersey and Dee) 
 since 1892. 
 
 It has in the last five years extended greatly about Bidston Junction among the sand . 
 There is much more of it north of Liverpool. In C. C. Babington's Flora it is given as 
 "American" and its home "Lancashire Sandhills." When newly opened the flowers are 
 very fragrant, but only last twenty-four hours. 
 
 Seeds inclosed, more of which I can send if desired; came from Bidston Junction. This 
 plant is identical with that of North Liverpool. 
 
 At Bidston, the characteristic sandhill plants among which it flourishes are Erysimum 
 chetraiithiodcs, Brassica vionensis, Anthyllis vulncraria, Psamma arcnaria, Equisetum max- 
 imiim, Lychnis alba, Senccio jacobaea, Salix arenaria. 
 
 It extends chiefly along the railway lines, which is suggestive as to its origin. 
 
 Inclosed enlargement from one-fourth plate shows the plant in situ, with flowers and 
 fruit stems in end, September, 1905. This year I saw the same type of this plant about the 
 railway at Llangollen in the northwest. 
 
 ^Ir. H. Stuart Thompson (Thompson, 1906) quotes Watson's New Botan- 
 ist's Guide (1837) as the first record of 0. biennis on the coast of Somerset and 
 says, "indeed it is quite a feature in the landscape." It was this sentence 
 that aroused interest in these special plants, as anyone conversant with the 
 weed-like aspect of the 0. biennis of this region would scarcely credit them 
 with anything like a spectacular appearance. 
 
 Hall in 1845 (Hall, 1845, p. 37) speaks of the plant as being extensively 
 naturalized on the Liverpool sandhills, and James Edward Smith in 1806 
 (Smith, 1806, p. 1534) figures a plant that was gathered on the extensive and 
 dreary sandbanks on the coast a few miles north of Liverpool, where millions 
 of the same species have been observed by Dr. Bostock and ]\Ir. John Shep- 
 herd, perfectly wild, covering a large tract between the first and second range 
 of sandhills. Some natural cause has no doubt established it here, though 
 possibly from the opposite side of the Atlantic. The Smith plate and the 
 plant figured by Baxter (Baxter, 1839, p. 257) certainly do not represent the 
 common North American 0. biennis, and the closing note in Baxter, referring 
 to the sudden expansion of the flowers, applies more readily to such large- 
 flowered species as 0. grandiflora,0. lamarckiana, and O.argillicola than to the 
 smaller-flowered ones, like 0. biennis, 0. muricata, and 0. oakesiana. 
 
8 MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 
 
 A recent writer on ths flora of Lancashire records that 0. biennis has been 
 estabhshed there for the last 70 or 80 years and that "whenever the land is 
 disturbed, or the sand removed to form new roads, this plant is one of the 
 earliest to grow upon it, and although its conspicuous flowers make it an easv 
 prey for constant plucking, it survives these depredations and continues to 
 spread more and more." 
 
 \\'ith Mr. Green's photograph and these references in mind, it was with no 
 little curiosity that the cultures from his seeds were watched. The seed- 
 lings raised from them earlv in 1906 proved to be indubitable O. lamarckiana. 
 In the seed-pan 2 seedlings of 0. lata and 4 seedlings of 0. rubn'ncrvis were also 
 recognized among a preponderating number of the O. lamarckiana. These 
 plants were observed unto maturity. 0. rubrinervis displayed all the char- 
 acters ascribed to it by Professor De Vries, with the exception that they were 
 not quite as large as some that had previously been studied in the New York 
 Botanical Garden. The two plants of 0. lata, on the contrary, were more 
 robust than any we had had in New York from Professor De Vries's seed, and 
 the pollen being more abundant than is usual in the species, efforts were made 
 to self-pollinate the flowers in the attempt to obtain pure seed. In due time 
 a few capsules ripened, containing a very small amount of seed, from which 
 were raised in December, 1906, 8 0. lamarckiana, 10 0. lata, 2 0. oblonqa, and 
 1 0. albida seedlings. 
 
 Before these facts were ascertained, the presumption was that the British 
 O. biennis was a slightly larger-flowered evening-primrose, and in our cul- 
 tures from foreign seed we usually named it "European biennis," of which 
 no real equivalent appears to be known here. Whether our native small- 
 flowered species is also known in Great Britain is not apparent from the man- 
 uals, as the species there recorded indicates a more ornamental plant than 
 the one with which we are familiar here, and ours is certainly not "adapted to 
 the shrubbery." 
 
 The misunderstanding in regard to the identity and provenience of O. 
 biennis, 0. lamarckiana, and 0. muricata seems to be vfidespread. There is 
 now no doubt as to the identity and original habitat of 0. grandiflora Alton. 
 The time and manner of its introduction into England are also known. It 
 escaped from gardens after it was taken into England in 1778, and is now 
 found growing wild in many places, according to unverified reports. 
 
 But as to the advent of lamarckiana we are still at a loss for positive proof. 
 The plant referred to by Miller, if correctly identified with this form, would 
 place it in Haarlem in 1757, 21 years before the discovery of grandiflora. 
 The next we hear of the grandiflora of Lamarck is in the Paris garden in 1797, 
 which was seen by Seringe not to be the grandiflora of Alton as known in 
 England and was renamed after its illustrious discoverer. Next we have the 
 striking form which is described above as appearing on the coast of Somerset 
 
MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 9 
 
 in 1837, and at other places since then, forming now quite a feature of the 
 Liverpool sandhills, which has been shown by our cultures to be the lamarck- 
 iana around which so much interest clusters. It would seem difficult in the 
 light of these facts to find any basis upon which the conclusion that this plant 
 had been derived from 0. grandi flora might rest. 
 
 The plant originally introduced into Europe under the name of 0. hicnnis 
 seems to have been a large-flowered form altogether different from the plant 
 under that name as known in America. Then O. lamarckiana grown from 
 imported seeds or their progeny, and O. qrandiflora from Alabama, present 
 a very diflferent aspect and have but little resemblance to the 0. biennis of 
 the waste lands of eastern America. 
 
 THE MUTABILITY OF OENOTHER.4 LAMARCKIANA. 
 
 About 2500 plantlets of 0. lamarckiana derived from various sources have 
 been brought under observation, as described above. Of these 106 or a little 
 over 4 per cent were mutants. One culture (C.1.2) appeared to give a much 
 higher proportion of mutants, but a failure to make a record of one lot of 
 discarded plants makes it impossible to confirm this. The behavior of this 
 lot of plants and also the confusion of fasciated specimens of 0. lamarckiana, 
 in the rosette stage, with mutants, led the senior author to assert that the pro- 
 portion of mutant derivatives in pedigreed cultures of O. lamarckiana had 
 been increased from 5 to 6 per cent of the progeny (MacDougal, 1905). 
 In no culture of later date has it been possible to exceed the maximum propor- 
 tion of 5 per cent, and the greater majority fall much below this figure. While 
 the coefficient of mutability has not been modified yet, one or two forms have 
 appeared which were not recognizable in a rosette stage by Professor De 
 Vries, and probably constitute an extension of the range of mutability. This, 
 however, may not be interpreted to mean an effect of local conditions. All 
 of the mutants now known did not occur in the cultures in Amsterdam 
 during the first few years of the cultures. Many, in fact, occur so rarely 
 that by the law of probability it is necessary to grow many thousands of speci- 
 mens in order to secure one individual of the rare type. In accordance wnth 
 this same principle it is to be seen that the spreading of the cultures of O. 
 lamarckiana may be expected to bring to light other rarer derivatives. When 
 such forms are found it is not to be interpreted as a result of local conditions 
 of the culture in which they appear, but simply in accordance wnth the coeffi- 
 cient of mutability of the parental type with, respect to the form in question. 
 
 The uniformly low frequency of mutants in material procured from seeds- 
 men is a matter not altogether understood. Among the probabilities to be 
 taken into account is that the seeds may have lain in storage for more than 
 one season before being planted, in consequence of which the unequal hardi- 
 ness of the mutants and the parent might operate to decrease the apparent 
 
lO MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 
 
 frequency of the derivatives. No proof has been produced to show that the 
 frequency of mutations might be increased beyond the Hmit of 5 per cent 
 found by De Vries, but it is a well-known fact that mutations of all kinds may 
 be decreased by inadequate nutrition, and it is quite possible that the parental 
 plants grown to furnish seeds did not have the care and nourishment of pedi- 
 greed individuals used in the cultures. The appearance of the rare ruhrinervis 
 in the culture might be due to a hybridization with that form, since no evi- 
 dence has been offered to show that the fertilizations were guarded. 
 
 The five cultures described together furnished albida, nanella, oblonga, lata, 
 gigas, scintillans, and nine other forms not positively identifiable at this time. 
 Seeds have been preserved, however, and it is hoped we may be able to 
 present descriptions of at least some of the forms at the close of another year. 
 
 Since these cultures were carried through, De Vries has published the 
 results of his cultures from seeds obtained from seed merchants. 2000 plantlets 
 from seeds obtained from Haage & Schmidt of Erfurt contained i specimen 
 of ruhrinervis, which arises but rarely in any culture, i of oblonga, and 3 of 
 nanella. 
 
 Two cultures were made from seeds furnished by Vilmorin and grown in 
 1 898- 1 899. The first gave 14 nanella, 3 lata, 3 scintillans, 1 albida, i oblonga, 
 and a few other divergent forms in 3500 seedlings. A second test yielded 
 3 lata, I nanella, and i ruhrinervis (?) in 600 seedlings. In both cases the pro- 
 portion of the mutants was not above 5 per cent, and was even less than that 
 described above (De Vries, 1905). 
 
 In view of the numerous observations described above upon material from 
 the most widely separated sources, which show mutability, suggestions that 
 the mutability of Lamarck's evening-primrose has been consequent upon 
 its segregation in the Amsterdam Botanical Garden may no longer be taken 
 seriously, although no doubt the well-worn phrase will be duly rehearsed 
 from time to time by careless critics. 0. lamarckiana may be of hybrid origin, 
 and what species may not be; but if so this origin is not recent, nor can 
 sufiicient time be found to hunt down all of the unprofitable alternatives that 
 are offered for consideration. 
 
 Exact records have already been made of lamarckiana, ruhrinervis, nanella, 
 and gigas in previous papers (MacDougal, Vail, Shull & Small, 1903 and 
 1904), and below are the principal taxonomic characters of albida, oblonga, 
 scintillans, hrevistylis, and lata, by which systematic descriptions of 9 of the 
 15 mutant derivatives iteratively yielded by lamarckiana are afforded from 
 material grown in New York (see plate i). 
 
 Note. — The illustrations of 0. lamarckiana and O. ruhrinervis prcvionsly published by 
 MacDougal (1903) were reduced without proper indication of the scale. 
 
 The capsule of 0. gigas is incorrectly given in another paper as 2 nun. in length when it 
 should be 2 cm. (MacDougal, Vail, Shull & Small, 1905.) 
 
PLATE 2. 
 
 ^=^^ 
 
 Oenothera scinlillans. 1, leaf from rosette; 2, leaf from stem; 3, bract; 4, bud with 
 bract; 5, flower with petals removed; 6, petal; 7, capsule. 
 
MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE) OENOTHERAS. II 
 
 OENOTHERA SCINTILLANS. 
 
 O. scintiUans was first seen to originate as a mutant in 1888 and has appeared 
 in this manner in 14 individuals at different times in De Vries's cultures. As 
 previously described in this paper, it has appeared in 4 specimens in the New- 
 York Botanical Garden during the present year. 
 
 O. scintiUans is to be noted as an ever-sporting form which in some strains 
 gives a progeny 60 or 70 per cent of which is composed of lata, oblonga, lamarck- 
 iana, and nanella, while in other strains but 30 per cent will be included in 
 these forms. The forms thus derived from scintiUans are in no wise different 
 in their hereditary qualities from the same types derived from other and pure 
 lineages. Cultures were made from seeds furnished by Professor De Vries 
 from a strain which was supposed to yield 60 to 70 per cent of the other types 
 named. 
 
 Seventy-eight individuals were obtained from the culture, 15 of which resem- 
 bled scintiUans, ^6 lamarckiana, 16 oblonga, and i was present which did not 
 reach maturity, but suggested some of the incomplete forms mentioned in ' ' Die 
 Mutationstheorie," ^rohahXy sublinearis. In the seedlings 3 months old, the 
 rosettes were 14 to 17 cm. in diameter, dense, and spreading flat on the ground ; 
 outer leaves, 6 to 7 cm. long, blades about 5 cm. long, 2.5 to" 3 cm. wide, 
 oblong-obovate, widest about the middle, obtuse or acutish at the apex, 
 tapering gradually to the white-margined petiole, blue-green, shining, puber- 
 ulent, with occasional small red spots above, glabrous or nearly so, and lighter 
 green beneath, somewhat brittle. At 6 months of age the rosettes were 
 dense, the leaves lanceolate, obscurely and minutely denticulate, green and 
 shining, puberulent on margin and sparingly on upper surface, midvein broad, 
 petioles variously winged. 
 
 The adult plant was short, with few branches which were fairly rigid and 
 ascending. The lower part of the stem appeared to be terete or nearly so, 
 while the upper portion was angled and channeled. The stem and branches 
 were hirsute with spreading hairs. 
 
 The stem-leaves were finely pubescent on both surfaces and m_uch crinkled 
 on the margins and midribs; irregularly toothed, oblong, 7 to 9 cm. long, 30 to 
 35 mm. wide, abruptly tapering at base, acutish at the apex. The petioles 
 were short and the entire organ was a deep shining green. 
 
 Conic portion of the bud 15 to 20 mm. long, finely pubescent, with short, 
 spreading hairs. Free tips 5 mm. long, erect. Hypanthium 30 mm. long, and 
 reflexed calyx-lobes 25 cm. long. Ovary 6 to 7 mm. long. 
 
 The petals are thin, 25 mm. long and 30 mm. in width, barely emarginate. 
 Filaments 15 to 18 mm. long, very slender and strongly upcurved. Anthers 
 7 to 8 mm. long. Pistils slightly longer than the stamens ; lobes of the stigma 
 very slender, 4 to 5 mm. long (plate 2). 
 
12 MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 
 
 OENOTHERA BREVISTYLIS. 
 
 A culture was made from seed harvested by De Vries in Amsterdam in 1 899 
 and said by him to contain about 25 per cent hrevistylis, the seed being the 
 product of the first generation of a cross between 0. lamarckiana and 0. 
 bvevistylis. Twenty-one plantlets appeared and were inspected, and some 
 seemed to form rosettes with much more rounded and abruptly pointed apices 
 than 0. lamarckiana. This was the only distinction that could be made 
 between the parental type and the derivative, during what would naturally 
 be the first year of development. In addition one example of 0. nanella as a 
 mutant from the parental type was included. With the beginning of the 
 formation of lengthened internodcs in the stem but little difference between 
 the plants was discernible until flower-buds appeared. At that time the 
 leaves on the terminal portions of the branches and the bracts seemed rela- 
 tively much broader, and the flower-buds might be distinguished at some 
 distance by the fact that they were more cylindrical than the parental type, 
 and were abruptly short-pointed. An examination of the structure of the 
 flower revealed the fact that the style is much shorter than in 0. lamarckiana 
 and that the stigma generally appears in the throat of the flower and some- 
 times is not to be seen without tearing the hypanthium apart. 
 
 It has been noted by De Vries that the length of the style varies widely 
 (as much as i cm.), and in this is offered an example confirmatory of results 
 communicated by the authors in a recent paper (MacDougal, Vail, Shull & 
 Small, 1905), in which the mutant characters were found to offer a wider 
 amplitude of variability than the correspondent characters of the parental 
 type. In addition to the above general characters, it is to be noted that, as a 
 consequence of the extreme shortness of the style, pollination fails in many 
 cases and comparatively few capsules are matured. 
 
 The evidence offered by this plant is of the greatest interest in connection 
 with questions concerning the survival of mutants in the habitat of the paren- 
 tal form, especially in view of the intemperate indulgence which many authors 
 show in theoretical discussions upon questions of this character. 
 
 0. hrevistylis was discovered August 20, 18S6, by De Vries, as represented 
 by 2 individuals. Despite the greatest care and the most rigid inspection it 
 has never been observed to arise in any of the cultures in Amsterdam or New 
 York, and the supposition is certainly allowable that it is no longer included 
 among the nmtants given off by the parental type. The characteristic qual- 
 ities of O. hrevistylis are recessive when hybridized with 0. lamarckiana, and 
 consequently the first generation resembles O. lamarckiana in outward form 
 but carries 0. hrevistylis, which reappears in the next or the second generation 
 of the hybrid, forming on an average one-fourth of the progeny, according to 
 the simple Mendelian formula. 0. hrevistylis is a retrogressive departure from 
 
PLATE 3. 
 
 Oenothera breris/ylis. 1, leaf from mature rosette; 2, bract and capsule; 3, stem-leaf; 
 4, flower with petals removed; 5, bud; 6, hypanthium split open, showing the short, 
 imperfect style; 7, petal; 8. bract; 9, capsule. 
 
PLATE 4. 
 
 2 4 
 
 Oenothera lata. 1, leaf from mature rosette; 2, bract and capsule; 3, bract; 4, bud and bract; 
 
 5, flower with petals taken off; 6, petal. 
 
MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 1 3 
 
 the parental type, matures comparatively few seeds, and is dominated by the 
 parental characteristics in crossing, and it has survived in its original habitat 
 and blooming specimens have been observed there for twenty years in com- 
 petition with the parental form. 
 
 The more important anatomical features of O. brevistylis were described 
 by Pohl (1895), but it seems desirable that its general taxonomic aspect, as 
 presented in tha cultures in New York, should be put on record. 
 
 The rosettes are open, leaves ovate-spatulate, crinkled, the upper ones 
 approximitely denticulate, bright, dark green, puberulent beneath, less so 
 above, conspicuous broad veins, petiole winged to near the base, much more 
 obtuse and rounded at apex than 0. lamarckiana. 
 
 The adult shoot resembles 0. lamarckiana very closely, even in the form 
 of the stem-leaves. The bracts, however, seem slightly thinner and appear 
 somewhat broader than those of the parental type. 
 
 The conic parts of the buds are 2 to 5 cm. long, appearing more cylindrical 
 than the parental form, and are covered with very short, spreading hairs. 
 The buds are distinctly tinged w4th red, and the free erect tips are unequal 
 and are not spreading. 
 
 The petals are 45 to 50 mm. long and 40 mm. wide, being firm and more or 
 less deeply emarginate. The filaments are 2 cm. long, the anthers i cm. long 
 and slender. The pistil does not usually emerge from the tube of the hypan- 
 thium, the stigma being seen in the very throat of the flower. The stigmatic 
 lobes are very irregularly developed. 
 
 But few capsules are matured. The capsules are 15 to 18 mm. long and 
 6 to 7 mm. in diameter at the base, being finely pubescent with a few spreading 
 hairs, bright green and slender, tapering slightly toward the apex. 
 
 The hypanthium in this species bears about the same relation to the reflexed 
 sepals as in the parental form (plate 3). 
 
 OENOTHERA LATA. 
 
 0. lata was first seen by De Vries in the plantlets produced by seeds har- 
 vested at Hilversum, Holland, in 1886, and brought into bloom in 1887, and 
 he has observed the origin of a total number of 493 as mutants in a progeny of 
 the parental type embracing 130,000 plants. 
 
 It has been observed to arise in New York from the third generation of 
 a strain grown from seeds furnished by Professor De Vries, and also in seeds 
 obtained from Vilmorin in Paris. 
 
 A sowing was made of seeds of 0. lata X 0. lamarckiana, furnished by Pro- 
 fessor De Vries. This form is characterized by atrophied stamens which 
 produce only a few pollen grains, of fairly normal structure, which seem to 
 be incapable of producing fertilization; but when the plant is pollinated by 
 O. lamarckiana, lata characters appear in the first generation as forming from 
 
14 MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 
 
 4 to 45 per cent of the progeny, or an average of 22 per cent. In the cultures 
 made in New York, lata, if present, was mistaken for lamarckiana in the seed- 
 hng stage and was discarded. Among the plants of this progeny which were 
 finally brought to bloom were albida, oblonga, nanella, and another type not 
 recognizable by the authors. 
 
 Laia is to be regarded as a retrogressive departure from the parental type, 
 and is most readily recognized by the rounded leaves of the rosettes, and these 
 organs become very much crinkled in later stages of development. The 
 leaves are of a deep green, and the flower-buds are much thicker than any 
 other evening- primrose known, except gigas, perhaps. 
 
 The principal taxonomic characters are as follows: The general habit of 
 this species resembles that of 0. gigas more nearly than any other of the 
 mutants, being sparingly branched, and having short, stoutish branches on 
 the upper portion of the stem. All of the branches are ascending, and in 
 many of the plants the terminal portion of the stem is bent over. 
 
 The stem is channeled and angled, brittle, and both stems and branches 
 are hirsute. The stem-leaves are finely pubescent, 10 to 15 cm. long, 4 to 5 
 cm. wide, numerous, and with the laminae remotely and shallowlv toothed. 
 The leaves of the rosettes, even in the very early stages, are noticeablv rounded 
 and obtuse at the apex, and those of the lower part of the stem are spatulate- 
 oblong and vary to ovate-oblong on the upper portions of the stem. Those 
 on the upper part of the stem are either obtuse or acutish at the apex and 
 taper to a margined petiole at the base. The leaves, even in voung rosettes, 
 are closely bunched into a head and are so heavily crinkled and thrown into 
 convexities and concavities between the veins as to be easily distinguishable 
 from all other forms in all stages of development. 
 
 The bracts are large, ovate to ovate-oblong, acute or obtuse at the apex, 
 cordate or subcordate, and clasping at the base. 
 
 The conic portions of the heavy buds are 2 cm. long and i cm. in diameter in 
 the basal portion, tapsring so slightly as to appear nearly cylindrical, and are 
 finely pubescent with spreading hairs. The free tips of the buds are spreading, 
 stout, 4 to 5 mm. long. The hypanthium is about 35 mm. long, stout, finely 
 pubescent and much longer than the reflexed sepals (plate 4 and plate 5, a). 
 
 The petals are thick, crinkled, and do not expand fully, being about 35 to 40 
 mm. long, and 45 to 50 mm. wide. The filaments are 18 to 20 mm. long, 
 while the anthers are imperfect, being very slender and about 8 mm. in length. 
 The microscopical examination of the anthers showed a few pollen grains 
 apparently perfect, yet no fertilization has ever been accomplished with them 
 in Amsterdam. 
 
 In a recent paper R. R. Gates (1907) describes the results of an investigation 
 of the pollen development of 0. lata in which he ascribes the failure of the 
 
PLATE 5. 
 
 # # 
 
 t^ 
 
 B 
 
 A. KosL-ttu otOciiotlKia l:il;i, fr)iir months old, anil separate leaves of the same age. 
 H. Rosette o( Oenothera alhida. tour months old, and sejiarate leaves of the same age. 
 
MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 1 5 
 
 pollen to other causes than the ingrowth of the tapetum to fill the loculus. 
 He found that development may proceed to the formation of tetrads, but 
 frequently degeneration begins in the resting stage of the first mitosis. The 
 successful pollinations recorded below, however, demonstrate the functional 
 maturitv of the pollen in many cases. Heterochromosomes were found to 
 arise in the prophase after synapsis, both in lata and in the lamarckiana hybrid 
 with lata. Inferential support was secured for the supposition that the 
 changes constituting mutation occur during the reduction divisions. 
 
 The pistil is barely longer than the stamens, with the stigmatic lobes, heavy 
 and club-shaped, 6 to 7 mm. long. The capsules are 13 mm. long, about 6 to 7 
 mm. in diameter in the thickest portion, finely pubescent, many angled, and 
 tapering slightly toward the apex. The terminal rosettes on the stem and 
 branches were small and regular. 
 
 While this form has hitherto been found incapable of producing mature 
 pollen capable of effecting fertilization in repeated trials made in Amsterdam 
 and New York, a notable exception is to be recorded. A package of seeds of 
 O. lamarckiana, which is growing wild near Birkenhead, England, was received 
 from Mr. C. Theodore Green in 1905. A sowing made from this lot yielded the 
 parental form, rubrinervis, and a few individuals of lata, which agreed with 
 this form as derived directly from pure cultures so far as all external charac- 
 ters were concerned. Many individuals previously examined in various 
 laboratories showed pollen which appeared normal under the microscope, yet 
 no fertilizations could be secured. In the present instance, however, pollen 
 appeared so plentiful that another attempt was made, the treated pistils 
 being carefully guarded from pollination from the parental or other forms. 
 A few pods were matured and some seeds were secured as a result. These 
 were sown as soon as ripe, in September, 1906. The progeny showed 10 lata, 
 80 lamarckiana, i albida, and 3 ohlonga. No material difference between this 
 and progenies resulting from pollination with lamarckiana was thus shown. 
 In the latter case the proportion of lata is often as low as 4 per cent of the 
 entire prog^nv and has not been recorded to have gone beyond 45 per cent. 
 
 OENOTHERA OBLONGA. 
 
 O. ohlonga was first observed by De Vries at Amsterdam in 1895, although 
 it doubtless appeared in his cultures previous to that time, as he suggests. In 
 all it has arrived in 700 mutants and in the pedigree-cultures described in this 
 paper in 35 individuals, only a part of which came into bloom in August, 1905, 
 agreeing with the observations in Amsterdam. Oblonga is one of the most 
 easily recognizable of the derivatives of 0. lamarckiana, although the distin- 
 guishing characters do not readily lend themselves to taxonomic description. 
 
 The rosettes are not very dense ; the leaves in the young rosettes are nar- 
 rowly ovate-lanceolate, rather thick and fleshy, with broad midveins which 
 
1 6 MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 
 
 are sometimes distinctly reddish. The laminae are shining green above, more 
 or less puberulent on both surfaces, and obscurely denticulate. 
 
 The adult plant is less than i meter in height and is very sparingly branched 
 below, with a few very short branches above. The basal branches are ascend- 
 ing and do not reach above half the height of the stem. The stems are 
 strongly channeled and angled. 
 
 The stem-leaves are crowded, hang down, and are pubescent on both sur- 
 faces, 7 to lo cm, long, oblong-elliptical, remotely and shallowly toothed, 
 acutish or obtuse at the apex, and irregularly narrowed into a short margined 
 petiole. The laminae are strongly crinkled, dark green, and become tinged 
 with crimson with age. 
 
 The bracts are oblong or ovate-oblong, acute at the apex, cordate or sub- 
 cordate at base. The buds are reddish, the conic portion being about 2 cm. 
 long and finely pubescent with short spreading hairs and a few longer ones. 
 The free erect tips are about 3 to 5 mm. in length. The petals are thin, 
 generally crinkled, somewhat emarginate and crenate, and are 3 cm. long and 
 4 wide. In the unfolding of the flower the petals open only so far as to form 
 a cup-shaped corolla. The hypanthium is about 3 cm. long, slightly longer 
 than the calyx-lobes. The slender ovary is about 8 mm. long. The filaments 
 are 15 mm. long and the anthers 8 mm. The pistils are slightly longer than 
 the stamens, the stigmatic lobes being about 5 mm. in length. 
 
 The capsules are 22 to 26 mm. long and 6 to 7 mm. in diameter at the thick- 
 est part, finely pubescent, slightly angled and shining green, tapering slightly 
 to the apex (plate 6). 
 
 OENOTHERA ALBIDA. 
 
 O. albida was first recognized by De Vries as a mutant in 1888 and was 
 brought into bloom in 1896. It is distinguishable even in the earlier stages 
 by the paler color of the leaves, which in the rosettes have upturned margins 
 and are variously twisted. Seventeen examples appeared in the cultures 
 previously described during 1905, of which 6 were mutants from . lamarckiana 
 as found introduced on Nantucket. The following taxonomic characters 
 were observed : 
 
 Seedling about 3 months old: Rosettes 15 to 17 cm. in diameter, thin, 
 somewhat raised above the slender rootstock; leaves 7 to 10 cm. long, 2.5 to 
 3 cm. wide, erect-spreading, the apex recurved and nearly touching the 
 ground; blades 5 to 7cm. long, oblong, broadest at about the middle, acutish 
 or mostly obtuse at the variously rounded apex, gradually tapering into the 
 almost translucent, margined petiole, pale light green, very thick and brittle, 
 margins undulate-denticulate, some of the 3^ounger leaves strongly wrinkled 
 on the margins. About 2 months later the rosettes were irregular, leaves 
 twisted and various-erected, convexed upwardly, pale yellowish-green, approx- 
 
PLATE 6. 
 
 Oenothera oblcnga. 1, leaves from mature rosette; 7. stem-leaf; 3, bracts; 4, petal; 5, capsule; 
 
 6, bud; 7, flower with petals removed. 
 
PLATE 7 
 
 Oenothera albida. 1, leaf of mature rosette; 2, stem-leaf; 3, bract; 4, bud; 5, flower with 
 
 petals removed; 6, petal; 7, capsule. 
 
MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 1 7 
 
 imately denticulate, oblong, soft-hairy on both surfaces, petioles winged half- 
 way to base. 
 
 The adult plant was seen to be irregularly branched, the branches often 
 exceeding the main stem in length, which never reached a height of a meter. 
 Stem zigzag, rather stout, channeled and angled, branching mostly at base, both 
 branches and stems very brittle. Branches and stems hirsute, with spreading 
 hairs and closely pubescent. Internodes short, with the leaves crowded. 
 
 Stem-leaves oblong to oblong-lanceolate, more or less regularly denticulate, 
 mostly obtuss or acutish. nearly sessile, tapering at base, finely pubescent 
 on both surfaces, very brittle and somewhat crinkled. 
 
 The bracts are oblong, acutish, or obtuse, slightly clasping, subcordate, 
 erected when young, reflexed when mature, margins upturned. 
 
 The buds are distinctly reddish, 3 cm. long, finely pubescent, with spreading 
 hairs and the free tips erect and 6 mm. long. 
 
 Hypanthium 4 cm. long, slightly longer than the sepals. Petals thin, 
 35 to 40 mm. long, and 40 to 50 mm. wide, reflexed from the middle, deeply 
 emarginate at the apex. Filaments 2 to 3 cm. long, anthers 12 mm. long. 
 Pistils longer than the stamens, with the stigmatic lobes 5 mm. in length. 
 
 Capsules 35 mm. long and 5 to 6 mm. in diameter at thickest portion, finely 
 pubescent and reddish, tapering slightly to apex. It is noticeable that com- 
 paratively few capsules are matured by this species. The terminal rosettes 
 are symmetrical and the buds erect and prominent (plate 5, b, and plate 7). 
 
 HYBRIDIZATION OF A MUTANT AND ITS PARENTAL TYPE. 
 
 In order to test the results which have been presented by Professor De Vries 
 with respect to the behavior of mutants in hybridization, a number of flowers 
 of 0. lamarckiana were castrated early in order to avoid entirely any proba- 
 bility of self-pollination, and these flowers were pollinated from unopened 
 buds of 0. ruhrinervis. The resulting capsules were harvested early in Sep- 
 tember, 1904, and sown in sterilized soil in the greenhouses in November, 1904. 
 Early in January the seedlings were large enough to make the separate types 
 discernible, and the task of discarding the duplicates was begun. The entire 
 progeny was found to include 90 of lamarckiana (and one of its mutants, of 
 which I was present) and 221 rubrinervis. The ruhrinervis, which is to be 
 regarded as a progressive mutant of latnarckiana, has been seen to surpass it in 
 vegetative vigor and rapidity of growth, besides producing seeds quite as 
 plentifully, and is sometimes dominant when crossed with the parent. Thus 
 De Vries found that when the above cross was made he obtained 19, 24, 68, 
 and 74 per cent of rubrinervis in four different lots, giving an average of 
 totals of 46 per cent rubrinervis produced in such crosses. The culture 
 described above yielded 71 per cent rubrinervis. 
 
THE FL^UCTUATIONS OF OENOTHERA LAMARCKIANA AND 
 
 ITS MUTANTS. 
 
 By George Harrison Shull. 
 
 The results of statistical studies made in 1904 on Oenothera lamarckiana Ser. 
 and two of its mutants, 0. nanella and O. rubrinervis, were so suggestive that 
 it was thought desirable to make a further attempt to trace the relations of 
 the mutants to their parental form, as indicated by the variations in some 
 of their differentiating characters. In order to reduce the probable error, it 
 was necassary to secure data from a larger number of individuals than the 
 cultures at the New York Botanical Garden could supply. Consequently cul- 
 tures of several of the Oenotheras were begun at the vStation for Experimental 
 Evolution. 
 
 The cultures which form the basis of this comparative study were the 
 following: 
 
 (a) About 90 spscimens of 0. lamarckiana were grown from seed supplied 
 under the designation D.1.8 by the senior author, being seed from plants 
 guarded and purely self-fertilized for three generations in the New York 
 Botanical Garden, and thirteen generations preceding in the Botanical Gar- 
 den at Amsterdam. In this, as in all the following cultures, no selection of 
 specimens was made, the desired number of plants having been taken consecu- 
 tively, beginning at one side of the seed-pan ; and in all cases the arrangement 
 in the garden was the same as that adopted at the New York Botanical Garden, 
 i. e., the plants were placed at a uniform distance of i meter from each other. 
 One of these 90 individuals was an O. lata and was omitted from the statistical 
 studies on this plot. The remaining 89 specimens were typical 0. lamarckiana. 
 
 (b) Ten rosettes of Oenothera lamarckiana were received by mail from Prof. 
 De Vries April 7, 1905, having been collected by him in the open field near 
 Hilversum, Holland, where over twenty years ago he secured the original 
 stock for his pedigree-cultures; 9 of these came to bloom and were uni- 
 form in appearance, agreeing well with the other cultures of O. lamarckiana 
 in all characters but those whose differences could be appropriately attributed 
 to the fact that these plants had grown as biennials, while the others had been 
 forced to annual bloom by giving an early start in the propagating house. 
 The chief differences between these plants and those of the culture described 
 above were seen in the greater height and more numerous and longer branches 
 of the plants from Hilversum. The rosette leaves were more numerous and 
 larger in both dimensions, but had the same form and the same degree of 
 crinkling. 
 
 18 
 
MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 1 9 
 
 (c) Eighty-eight specimens were grown from seed of an O. rubrinervis 
 which had appeared as a newly arisen mutant among the first generation 
 hybrid progeny of O. laniarckiana X biennis discussed in our earher account 
 (Mutants and Hybrids, p. 18). The occurrence of this mutant along with 
 several individuals of 0. lamarckiana was there attributed to an accidental 
 self-fertilization of the pistil parent, but there is no sufficient reason perhaps 
 for ruling out the occurrence of monolepsis producing Millardet hvbrids in 
 which the maternal characters alone appear. Whether this specimen of O. 
 rubrinervis was the product of self-fertilization or of monolepsis has no especial 
 significance for this study, since, whatever may have been its origin, it was a 
 true O. rubrinervis, though having no rubrinervis individuals in its ancestry, 
 at least for many generations. 
 
 The pollination of this specimen was left entirely uncontrolled, and as it 
 grew in close contiguity with other species of Oenothera, where insects were 
 busily engaged carrying from flower to flower the pollen of a dozen or more 
 different species, the progeny might have been expected to show a confusion 
 of hybrid forms of unknown affinities. It was a matter of some surprise, 
 therefore, that only i specimen in the 88 appeared to be a hybrid. The char- 
 acters of this obvious hybrid indicated that it was probably produced by pollen 
 from one of the small-flowered species, and it was considered by the senior 
 author to be nearlv if not quite identical with his 0. lamarckiana X biennis, 
 No. 2.32, though no definite analysis of its characters was made. The other 
 87 individuals consisted of 80 0. rubrinervis and 7 which were looked upon as 
 typical O. lamarckiana, although, as will be seen later, the results of the statis- 
 tical studies present some interesting exceptional features. 
 
 (rf) Twenty-seven individuals of O. rubrinervis were grown from guarded 
 and self-fertilized seed of one which arose as a mutant at Amsterdam in 1895 
 in a pure 0. lamarckiana pedigree guarded for six generations. 
 
 (e) Ninety-eight specimens of Oenothera gigas were grown from pure-bred 
 seed received fromProf. DeVries, the pedigree being known and fully guarded 
 for five generations from the time it appeared as a mutant in 1895. Unfortu- 
 nately onlv 4 of these came to bloom the first year, so that the studies here 
 given on Oenothera gigas must be considered wholly inadequate. 
 
 (/) Another lot of plants, 44 in number, were the offspring of Oenothera 
 lata which had grown surrounded by O. gigas in an isolated portion of the 
 Botanical Garden of Amsterdam. On the side of the pistil-parent the pedigree 
 of this lata had been continued for five generations from the time it arose as 
 a mutant, and in most of these generations 0. lamarckiana was the pollen par- 
 ent. As 0. lata is nearly or quite incapable of self-fertihzation, owing to its 
 lack of viable pollen, these plants must have been hybrids between 0. lata 
 and O. gigas, and as they were wholly unselected they formed a motley group. 
 Among them Oenothera lata was the most frequent, 17 in 44 belonging to this 
 
20 MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 
 
 species; 15 were 0. lamarckiana, and 2 were 0. nanella. The remaining 10 
 were not positively identified, though several which were at first thought to 
 be 0. lata but which were afterwards seen to differ from that species in certain 
 bud characters, especially in the possession of pollen-bearing anthers, may 
 have been true hybrids between 0. lata and 0. gigas. 
 
 The advantages of bringing together to one place for use in this line of 
 investigation material having such different past histories will be obvious. If 
 the various types of structure which have been separated and described under 
 specific names are not perfectly natural centers of stability, or rather of equi- 
 libration, but are the product of selection, it is inconceivable that the forms 
 resulting from selection by different persons for different periods of time and in 
 different environments should produce identical forms, and especially that 
 there should be identity in characters that were not taken into account and 
 could not well be taken into account by those making the selections. 
 
 Some differences are naturally to be expected in the characters of these 
 several lots of plants of the same species coming from different sources, 
 because of the known behavior of purely fluctuating characters, and the degree 
 of these differences will be contemplated with some interest. On the other 
 hand, the various sources of the material will make the comparisons between 
 the different species less satisfactory than might be wished. 
 
 A part of the advantage gained in the way of larger quantities of material 
 by bringing these cultures to the Station for Experimental Evolution is offset 
 by the fact that differences in soil, climate, and methods of culture from those 
 to which the cultures were subjected at the New York Botanical Garden in 
 1904 will make impossible the drawing of any conclusions this year regarding 
 the problem which originally incited these investigations, namely, the question 
 as to the presence or absence of regression in the mutants. As a partial 
 measure of the difference in conditions at the New York Botanical Garden in 
 1904 and at the Station for Experimental Evolution in 1905, a comparison of 
 the heights of the 0. lamarckiana used for the studies in the former year may 
 be made with those of the first lot (a) described on page 18. Both of these lots 
 were raised as annuals and should therefore give a reasonably fair indication 
 of the relative value of the resultant of all the factors affecting the fluctuating 
 characters of the two lots of plants. 
 
 The height-curves reduced to equal areas are shown superposed in fig. i, 
 and the principal constants are as follows : 
 
 New York Botanical Garden, 1004 
 
 Station for Experimental Evolution, 1905. 
 
 Mean. 
 
 Standard 
 deviation. 
 
 Centimeters. I CenHmeiers. 
 88.68 ±o..S5 ! 4-76±o.3q 
 86.98 ±0.69 8.37±o.j.9 
 
 Coefficient 
 of variation. 
 
 Per cent. 
 5.37 ±0.44 
 Q.62 ±0.57 
 
MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 21 
 
 It will be noted that there has been a slight fall in mean heights, but the 
 most noteworthy thing in this comparison is the great increase in the varia- 
 bility. A study of fig. I shows the difference in variabilit)' quite clearly in 
 the greater amplitude and less height of the 1905 curve. The interesting 
 fact becomes apparent that although the mean value of stem-heights in 1905 
 was less than that for 1904, some of the 1905 plants were taller than the tallest 
 in 1904. The explanation of these facts is very simple. The garden at the 
 Station for Experimental Evolution has been under cultivation so short a 
 
 *»-o 
 
 
 
 
 
 \ 
 
 1 
 
 ^ 
 \ 
 
 
 
 
 40 
 35 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 1 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 30 
 
 
 
 
 
 
 f 
 t 
 1 
 
 
 
 
 
 
 
 
 
 
 1 
 1 
 
 \ 
 \ 
 
 
 
 
 25 
 
 
 
 
 
 
 1 
 1 
 1 
 
 
 
 
 
 
 
 
 
 
 
 I 
 
 
 
 
 20 
 
 
 
 
 
 / f 
 
 
 V 
 
 \ 
 1 
 
 \ 
 
 
 
 
 
 
 
 ^/c 
 
 / 1 
 
 
 1 
 1 
 I 
 
 
 
 
 15 
 
 
 
 J 
 
 ^ 1 
 
 
 
 ^ 1-<- 
 
 
 
 
 
 
 } 
 
 
 
 
 
 ( 
 
 
 
 
 
 
 / 
 
 / 
 
 
 
 
 \ 
 
 
 
 
 
 
 / 
 
 / 
 / 
 
 
 
 
 \ 
 
 
 
 10 
 
 5 
 
 
 
 / 
 
 / 
 
 
 
 
 \ 
 
 
 
 
 / 
 
 s ; 
 
 / 
 
 / 
 / 
 
 / 
 
 ! 
 
 
 
 < 
 
 \ \ 
 
 
 
 
 
 / 
 
 / 
 
 / 
 / 
 / 
 / 
 
 
 
 
 
 \ \ 
 \ \ 
 \ \ 
 \ ] 
 
 \ 
 
 
 .^ 
 
 n 
 
 
 66-70 71-75 76-80 81-85 86-90 91-95 96-100 101-105 106-110 
 
 Fig. 1. — Variation in height of stem of Oenothera lamarckiana. Broken curve represents 
 the condition at the Xew York Botanical Garden in 1904, continuous curve that at the 
 Station for Experimental Evolution in 1905. 
 
 time that it still lacks much of that homogeneity that is desirable for work of 
 this nature. One corner of this plot of Oenothera lamarckiana dipped into a 
 slight depression which received the drainage from the roof of the laboratory 
 and was therefore kept, during a portion of the growing season, too wet for 
 the best development of the plants. A few specimens on this corner did not 
 make flower-stems at all, and it was to be expected that those nearest this 
 corner of the plot would be somewhat influenced by the same unfavorable 
 conditions and made to produce flower-stems of a shghtly less altitude. 
 
 The stature of Oenothera rubrinervis. — In 1904 a comparison was made 
 between the statures of Oenothera nanella and 0. lamarckiana, and it was a 
 
22 MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 
 
 part of the plan for the present study that this comparison should be contin- 
 ued, but unfortunately the seed of O. nanella gave only 3 germinations and 
 the continuation of this comparison must be postponed until another season, 
 but meanwhile it will be particularly interesting to make another comparison 
 of heights which will present something of a contrast with that between O. 
 nanella and its parent. It will be recalled that in that comparison the co- 
 existence of two striking features was taken to be more than a mere coinci- 
 dence, namelv, the great departure of the mean value in 0. nanella from the 
 
 30 
 
 £5 
 
 20 
 
 15 
 
 10 
 
 
 
 
 
 
 . 
 
 ^ 
 
 
 
 
 
 
 
 
 
 
 / 
 / 
 
 / 
 
 \ / 
 
 \ 
 
 
 
 
 
 
 
 
 
 /< 
 
 A 
 
 A \ 
 
 
 
 
 
 
 
 
 > 
 
 1 
 
 
 \ 
 
 / 
 
 
 \ \ 
 
 
 
 
 
 
 ^( 
 
 
 
 ^ 
 
 
 
 \ V 
 
 \ 
 
 \ 
 
 "> 
 
 \ 
 
 
 
 A 
 
 )'' 
 
 / 
 
 
 
 
 
 \ 
 
 \ 
 
 c 
 
 ) 0- ^ \ 
 
 
 ^ 
 
 65-70 71-75 76-80 81-85 86-90 91-95 96-100 101-105 106-110 III-II5 116-120 
 
 Fig. 2. — Comparison of the variability of stem-heights of Oenothera lamarckiana (broken curve) 
 and O. rubrinervis, as grown at the Station for Experimental Evolution in 1905. 
 
 mean of its parent, and a very great increase in the coefficient of variability 
 in the heights of the mutant. In comparing the heights of 0. rubrinervis with 
 those of O. lamarckiana there is seen a very much less marked departure from 
 the parental condition in the mean values, and a correspondingly slight 
 increase in the coefficient of variability. This comparison is shown graphic- 
 ally in fig. 2 and is expressed numerically as follows: 
 
 Variation in heights, 1905. 
 
 
 Mean. 
 
 Standard 
 deviation. 
 
 Coefficient 
 of variation. 
 
 Oenothera lamarckiana 
 
 Oenothera rubrinervis 
 
 Centimeters. 
 86.98 ±0.69 
 92. 18 ±0.99 
 
 Centimeters. 
 8.37 ±0.49 
 9.86 ±0. 70 
 
 Per cent. 
 
 9.62 ±0.57 
 
 10. 70 ±0.78 
 
 This may still be merely a coincidence, but two such coincidences are very 
 much less likely to occur than one, and the attempt will be made in the study 
 of other characters to trace the suggested parallel between the departure of 
 the mean from the parental condition and the increase in the value of the 
 coefficient of variabilitv. 
 
MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 23 
 
 The branching habit of Oenothera rubrinervis. — In discussing the branching 
 of Oenothera rubrinervis in 1904 the statement was made that "such a char- 
 acter as this is too largely influenced by the individual physiological vigor to 
 be of any value as a diagnostic character, " and this statement receives striking 
 confirmation in the conditions observed in 1905. The specimens of 0. rubri- 
 nervis grown at the Station for Experimental Evolution in the latter year were 
 totally different in their branching from those studied the year before at the 
 
 t-. t/ cd o>^ o oj 34-36 37-39 40-42 43-45 46-48 
 
 Fig. 3. — Comparison of the number of branches of Oenothera lamarckiana (shaded with lines rising 
 toward the right) and O. rubrinervis, a.s grown at the Station for Experimental Evolution in 1905. 
 
 New York Botanical Garden. There was in 1905 almost a complete sup- 
 pression of the large basal branches which were such a marked feature of the 
 specimens earlier described. Instead of finding Oenothera lamarckiana and 
 0. rubrinervis entirely distinct and separated by a wide gap in respect to 
 number of lateral branches, the curves in figure 3 show how greatly they 
 overlap. Their numerical relations may be studied in the following table : 
 
 
 Range. 
 
 Mean. 
 
 Standard 
 deviation. 
 
 Coefficient of 
 variation. 
 
 Number of the lateral 
 
 
 
 
 
 branches : 
 
 
 
 
 Per cent. 
 
 Oenothera lamarckiana . . 
 
 18 to 39 
 
 26.20 ± o.2>?, 
 
 3-99 ± 0.33 
 
 15-23 ± 91 
 
 Oenothera rubrinervis . . . 
 
 19 to 47 
 
 30 18 ± 0.72 
 
 6.12 ± 0.51 
 
 20.29 ± 1 .76 
 
 Total branch length in 
 
 
 
 
 
 decimeters: 
 
 
 
 
 
 Oenothera lamarckiana . . 
 
 40 to 150 
 
 91.40 ± I 49 
 
 17 97 ± 1-05 
 
 19.66 it 1 .20 
 
 Oenothera rubrinervis . . . 
 
 4T to 174 
 
 87.22 ± 2 36 
 
 23.46 ±1.67 
 
 26.89 ± 2-05 
 
24 MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 
 
 With respect to the number of branches, the figures in this table do not 
 fully represent the tendency in 0. rubrinervis to develop a large proportion of 
 its axillary buds into branches, as many of these buds had started into growth, 
 but for some reason had been arrested before they had attained a length of 
 I cm., under which length none were counted. Comparing these results with 
 those of 1904 it is found that with respect to vanahility Oenothera lamarck- 
 iana presents essentially the same condition in the two years, though the 
 values of both number and length of branches has increased in the latter year. 
 In Oenothera rubrinervis, on the other hand, although there has been in 1905 
 a great reduction both in number and length of branches and the coefficients 
 of variability are considerably changed, they are still much higher than in the 
 parent species. This will be best appreciated by a comparison of the results 
 for the two years as shown in the following table : 
 
 Variation in — • 
 
 Years. 
 
 Range. 
 
 Mean . 
 
 Standard 
 deviation. 
 
 Coefficient 
 of variation. 
 
 
 
 
 
 Per cent. 
 
 Number of branches of ( 
 
 1904 
 
 10 to 27 
 
 20. 27 ± 0.49 
 
 324 ± 0.35 
 
 15.70 lb 1.70 
 
 0. lamarckiana. { 
 
 1905 
 
 18 to 39 
 
 26.20 ± 0.33 
 
 .3 
 
 99 ± 0.33 
 
 15-23 ± 0.91 
 
 Number of branches of < 
 
 1904 
 
 34 to 63 
 
 42.35 ± 1.04 
 
 6 
 
 34 ± 0.73 
 
 1500 ± 1.70 
 
 0. rubrinervis. ( 
 
 1905. 
 
 19 to 47 
 
 30. 18 ± 0. 72 
 
 6 
 
 12 ± 0.51 
 
 20.29 it 1 .76 
 
 Branch-length of 0. J 
 
 1904 
 
 36 to 85 
 
 66.8 ± 2.0 
 
 13 
 
 5 ± 1.4 
 
 20. 2 ±2.2 
 
 lamarckiana. ( 
 
 1905 
 
 40 to 150 
 
 914 ± 1-5 
 
 18 
 
 ±1.1 
 
 20.0 ±1.2 
 
 Branch-length of 0. _< 
 
 1904 
 
 78 to 300 
 
 181. 9 ± 13.0 
 
 79 
 
 5 ± 9 2 
 
 43-7 ±5-1 
 
 rubrinervis. ( 
 
 1905 
 
 41 to 174 
 
 87.2 ± 2.4 
 
 23 
 
 5 ± 1.7 
 
 26.9 ±2.0 
 
 The greater variability of the mutants was taken in 1904 to indicate that 
 these forms might not yet have acquired, perhaps, that fixity of habit exhib- 
 ited by an older species, and the fact that Oenothera rubrinervis underwent 
 greater change in its mode of branching on being cultivated in a new environ- 
 ment than was exhibited by 0. lamarckiana when subjected to the same 
 change would accord well with this suggestion. 
 
 The increase in the number and length of the branches in 0. lamarckiana 
 and the considerable decrease in those of O. rubrinervis must not be inter- 
 preted, however, as in any sense an approach to actual identity in these two 
 species, in regard to branching habit, for in the latter year they were even 
 more distinct in general type of branching than before. Oenothera rubri- 
 nervis had simply lost its long basal branches, while 0. lamarckiana had retained 
 them. 
 
 THE SIZE AND SHAPE OF LEAVES OF OENOTHERA RUBRINERVIS. 
 
 Some of the difficulties attending the statistical study of leaves were pointed 
 out in our earlier paper, and the complete suppression of the basal branches of 
 O. rubrinervis in 1905 has made the continuation of the comparison between 
 
MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 25 
 
 the leaves of that species and those of its 
 parent of Httle value except as a demonstra- 
 tion of the difficulties which were to be 
 overcome. It seemed fair in the former year 
 to consider the basal branches of these two 
 species homologous, and leaves for the study 95 
 in 1904, being taken from about the middle 
 of those branches, were consequently as 
 nearly comparable as possible. In 1905 the 
 leaves of 0. lamarckiana were again taken 
 from the middle of the basal branches and 
 are thus comparable with those of 1904, 
 but in 0. rubrinervis the leaves were taken 
 from the middle of the lowest branches pres- 
 ent, and as these were not quite basal, the 
 leaves were not homologous with those of 
 O. lamarckiana. The first striking result 
 of this lack of homology is seen in fig. 4, 
 where the variation in leaf-lengths is com- 
 pared. The corresponding curves for 1904 
 showed almost exact identity in the length 
 of the leaves, while in this figure it is seen 
 that the length of most of the O. rubri- 
 nervis leaves corresponded in 1905 with the 
 lower part of the range of 0. lamarckiana. 
 When figure 5, representing the variation in 
 leaf-width, is compared with the corres- 
 ponding figure for 1904 it is seen that with 
 respect to this character the lack of homol- 
 ogy has had but little modifying effect, 
 and there is close resemblance between the 
 sets of curves representing the variation in 
 leaf-width for the two years, though both 
 species show a slight increase in the coef- 
 ficient of variability. The considerable 
 difference in respect to leaf-length, taken 
 in conjunction with essential identity in 
 leaf- width in 1905, has the effect to change 
 the ratios between length and width in 
 about the same degree as the change in 
 length, but in the opposite direction. Thus 
 the change in length was away from identity, 
 
 IIO 
 
 105 
 
 100 
 
 90 
 
 85 
 
 80 
 
 75 
 
 70 
 
 65 
 
 60 
 
 55 
 
 50 
 
 45 
 
 35 
 
 30 
 
 25 
 
 20 
 
 15 
 
 10 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ( 1 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 <1 
 
 ?f\ 
 
 
 
 
 
 ''\\° '"' 
 
 
 
 
 
 '1 '' ! 
 ' , ', 
 
 
 
 
 
 
 1 " ( 
 f ° 
 
 > ', 
 
 
 
 
 
 6 
 
 
 
 
 
 1 
 
 
 
 
 II 
 
 
 
 
 
 T ' 
 
 
 1 
 
 
 
 
 
 \ 
 
 c 
 
 
 
 
 ) 
 
 i 
 
 
 
 J4 
 
 
 I 
 
 
 A 
 
 4-6 
 50 
 
 71 
 75 
 
 96 
 100 
 
 121 
 125 
 
 14-6 
 150 
 
 Fig. 4. — Comparison of leaf -length of Oeno- 
 thera lamarckiana (broken curve) and O. 
 rubrinervis. 
 
26 MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 
 
 while that in the leaf 
 ratios was toward iden- 
 tity, as will be readily 
 seen on referring to fig. 
 6, which shows a much 
 more complete over- 
 lapping than do the 
 corresponding curves 
 for 1904. In other 
 words, since this ratio 
 was taken as a measure 
 of leaf-form, the form 
 of the leaves of O. 
 rubrinervis in the latter 
 year was but slightly 
 different from that of 
 the parent species, ex- 
 cept in those minor 
 details which were 
 pointed out in the 
 earlier study as pre- 
 venting the ratio be- 
 tween length and 
 width from being a 
 correct measure of 
 form. The lack of 
 homology in the leaves 
 measured in the two 
 species takes away all 
 significance of these 
 facts in their bearing 
 on the question of the 
 regression of 0. rubrinervis. The constants of the several curves representing 
 the variation in leaf measurements may be compared in the following table : 
 
 Fig. 5.- 
 
 -Comparison of leaf -width of Oenothera lamarckiana (broken curve) 
 and O. rubrinervis. 
 
 Variation in length of leaf: 
 
 Oenothera lamarckiana 
 
 Oenothera rubrinervis 
 
 Variation in width of leaf: 
 
 Oenothera lamarckiana 
 
 Oenothera rubrinervis 
 
 Variation in the ratio between width and 
 length : 
 
 Oenothera lamarckiana 
 
 Oenothera rubrinervis 
 
 Mean. 
 
 
 Millimeters. 
 88.91 ±0.35 
 70.96 ±0.24 
 
 36.56±o. 
 27.3o±o. 
 
 091 
 079 
 
 Per cent 
 42.27 ±0 
 39-34 ±0 
 
 ! 
 10 t 
 13 
 
 Standard 
 deviation. 
 
 Millimeters. 
 
 18.41 ±0.25 
 
 11.42 ±0.17 
 
 4. 7q6 ±o.o6j 
 3.687 ±0.056 
 
 Per cent. 
 5.411 ±0.072 
 5.888 ±o.o8q 
 
 Coefficient 
 of variation. 
 
 Per cent. 
 20.71 ±0.29 
 16. io±o.24 
 
 13. i2±o. 18 
 
 13 50±0.2I 
 
 1 2 . 80 ± o. 1 7 
 i4-97±o.23 
 
MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 27 
 
 The original measurements upon which the above figures and conclusions 
 are based are presented in the form of correlation tables in figs. 7 and 8. A 
 comparison of the tables with the corresponding figures for 1904 shows that 
 those representing 0. lamarckiana are very much alike in the two years, the 
 degree of correlation being even higher in the latter year than before, due in 
 part possibly to the slightly greater heterogeneity already pointed out in the 
 cultures at the Station for Experimental Evolution. On the other hand, in 
 Oenothera rubrinervis the corresponding tables for the two years are entirely 
 50r 
 
 Fig. 6. — Comparison of the ratio between leaf-length and leaf-width in Oenothera 
 lamarckiana (broken curve) and O. rtibrinerzis. 
 
 different, as would be expected from what has already been said. In this 
 species the coefficient of correlation has been considerably lowered. 
 
 The coefficients of correlation for the two species for the two years may 
 be compared in the following table : 
 
 
 1904. 
 
 1905. 
 
 Oenothera lamarckiana 
 
 0.7916 ± 0.0090 
 0.6604 ± 0.0119 
 
 0.8330 ± 0.0058 
 0.4452 ± 0.0171 
 
 Oenothera rubrinervis 
 
28 MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 
 
 Variation in bud characters. — It is a well-known fact that floral characters 
 fluctuate less in response to change of environment than vegetative characters, 
 though I know of no exact studies showing to what extent this is true. The 
 comparatively larger flowers of alpine regions find an explanation on this 
 basis since the alpine plants have generally migrated from the lowlands. 
 I have seen a specimen of Lobelia syphilitica less than 8 cm. tall and dwarfed 
 
 
 
 
 Xo 
 
 -9 
 
 -8-7 
 
 -6 
 
 — 5 
 
 -4 
 
 -3 
 
 -o 
 
 -i 
 
 
 
 1 
 
 2 
 
 3 
 
 4 
 
 5 
 
 G 7 
 
 8 
 
 9 
 
 10 
 
 11 
 
 12 
 
 13 
 
 
 
 
 
 
 
 
 10 
 
 15 50 
 
 55 
 
 60 
 
 65 
 
 70 
 
 75 
 
 80 
 
 85 
 
 90 
 
 95 
 
 100 
 
 105 
 
 110115!l2o'l25 
 
 130'l35 
 
 140 
 
 145 
 
 150 
 
 
 
 
 
 
 
 
 45 
 
 50 55 
 
 60 
 
 65 
 
 70 
 
 75 
 
 80 
 
 85 
 
 90 
 
 95 
 
 100105 
 
 110 
 
 115120125 
 
 130135IU0 
 
 145 
 
 150 
 
 155 
 
 
 
 
 
 Xi 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 - 
 
 -13 
 
 23 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 - 
 
 -12 
 
 24 
 
 
 1 
 
 
 1 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 3 
 
 
 - 
 
 -11 
 
 25 
 
 
 
 
 
 3 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 3 
 
 
 - 
 
 -10 
 
 26 
 
 
 
 
 o 
 
 2 
 
 3 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 8 
 
 
 
 -9 
 
 27 
 
 
 
 
 1 
 
 2 
 
 1 
 
 3 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 7 
 
 
 
 -8 
 
 28 
 
 
 
 1 
 
 
 10 
 
 10 
 
 7 
 
 3 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 31 
 
 
 
 -7 
 
 2'J 
 
 
 
 
 
 3 
 
 7 
 
 7 
 
 9 
 
 4 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 31 
 
 
 
 -G 
 
 30 
 
 
 
 
 2 
 
 4 
 
 8 
 
 10 
 
 9 
 
 9 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 13 
 
 
 
 -5 
 
 31 
 
 
 
 
 1 
 
 
 / 
 
 13 
 
 9 
 
 5 
 
 3 
 
 3 
 
 3 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 11 
 
 
 
 -4 
 
 32 
 
 
 
 
 
 5 
 
 12 
 
 14 
 
 27 
 
 G 
 
 8 
 
 6 
 
 3 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 81 
 
 
 
 -3 
 
 33 
 
 
 
 
 
 2 
 
 10 
 
 12 
 
 21 
 
 13 
 
 19 
 
 12 
 
 C 
 
 1 
 
 
 2 
 
 
 
 
 
 
 
 
 
 
 
 98 
 
 
 
 —2 
 
 34 
 
 
 
 
 
 
 
 
 12 
 
 20 
 
 22 
 
 19 
 
 10 
 
 G 
 
 4 
 
 5 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 105 
 
 
 
 -1 
 
 35 
 
 
 
 
 
 1 
 
 4 
 
 3 
 
 IG 
 
 10 
 
 27 
 
 18 
 
 13 
 
 10 
 
 4 
 
 2 
 
 
 
 
 
 
 
 
 
 
 
 108 
 
 
 
 
 
 36 
 
 
 
 
 
 
 G 
 
 3 
 
 5 
 
 8 
 
 14 
 
 10 
 
 181 a 
 
 5 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 79 
 
 
 
 1 
 
 37 
 
 
 
 
 
 
 
 G 
 
 11 
 
 7 
 
 14 
 
 21 
 
 20 
 
 10 
 
 8 
 
 G 
 
 2 
 
 
 
 
 
 
 
 
 
 
 105 
 
 
 
 2 
 
 38 
 
 
 
 
 
 
 
 2 
 
 3 
 
 G 
 
 9 
 
 12 
 
 26 
 
 11 
 
 21 
 
 4 
 
 4 
 
 6 
 
 1 
 
 
 
 
 
 
 
 
 105 
 
 
 3 
 
 39 
 
 
 
 
 
 
 
 2 
 
 3 
 
 1 
 
 3 
 
 
 
 8 
 
 iG 
 
 18 
 
 10 
 
 3 
 
 2 
 
 3 
 
 
 
 
 
 
 
 
 75 
 
 
 
 4 
 
 40 
 
 
 
 
 
 
 1 
 
 
 2 
 
 2 
 
 1 
 
 5 
 
 17 
 
 15 
 
 15 
 
 14 
 
 5 
 
 3 
 
 5 
 
 
 
 
 
 
 
 
 85 
 
 
 
 5 
 
 41 
 
 
 
 
 
 
 
 
 2 
 
 
 2 
 
 4 
 
 4 
 
 7 
 
 8 
 
 12 
 
 8 
 
 3 
 
 3 
 
 
 
 
 
 
 
 
 53 
 
 
 
 6 
 
 42 
 
 
 
 
 
 
 
 
 
 1 
 
 1 
 
 1 
 
 3 
 
 10 
 
 11 
 
 11 
 
 7 
 
 11 
 
 1 
 
 2 
 
 1 
 
 
 
 
 
 
 CO 
 
 
 
 7 
 
 43 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 2 
 
 4 
 
 4 
 
 7 
 
 4 
 
 11 
 
 6 
 
 2 
 
 
 1 
 
 1 
 
 
 
 
 43 
 
 
 
 8 
 
 44 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 3 
 
 9 
 
 9 
 
 4 
 
 1 
 
 
 
 
 
 
 
 27 
 
 
 
 9 
 
 45 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 2 
 
 4 
 
 10 
 
 5 
 
 4 
 
 1 
 
 1 
 
 2 
 
 
 
 
 
 29 
 
 
 
 10 
 
 46 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 ^ 
 
 3 
 
 3 
 
 2 
 
 9 
 
 
 1 
 
 
 
 
 16 
 
 
 
 11 
 
 47 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 O 
 
 3 
 
 
 O 
 
 2 
 
 1 
 
 
 
 
 1 
 
 
 12 
 
 
 
 12 
 
 48 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 1 
 
 
 
 1 
 
 
 1 
 
 
 
 
 
 
 1 
 
 
 
 13 
 
 49 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 3 
 
 
 
 1 
 
 
 1 
 
 
 
 
 
 5 
 
 
 
 14 
 
 50 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 o 
 
 1 
 
 1 
 
 o 
 
 
 
 1 
 
 
 7 
 
 
 
 15 
 
 51 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 1 
 
 
 
 
 
 
 2 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 2 7 
 
 33 
 
 75 
 
 91 
 
 111 
 
 91 
 
 121 
 
 110129 
 
 98 
 
 103 
 
 SI 
 
 63 
 
 53 
 
 35 
 
 12 
 
 8 
 
 6 
 
 2 
 
 
 
 2 
 
 1270 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Fig. 7. — Correlation between width (subject) andlength oi\eaV\nOcnothi:ralama)ckiana. 
 Measurements in millimeters. Coefficient of correlation, 0.8330 i 0.0058. 
 
 specimens of numerous other species bearing flowers almost as large as is 
 normal for the species. 
 
 Having sufficiently demonstrated the unsatisfactory character of height, 
 branching, and leaf-form for statistical comparisons because of their great sen- 
 sitiveness to even shght changes in the environment, I have decided that in the 
 buds is to be found probably the most satisfactory material for a continuation 
 
PLATE 8. 
 
 Huds of ()enotlier;i himarckiana and four of its mutants, showing: characteristic differences and 
 variability, ist row at top, Oenothera lamarckiana; 2nd row, Oenothera ruhrinervis; 3rd 
 row, Oenothera lata ; 4th row, left, Oenothera gigas, right, Oenothera nanella. 
 
5i 
 
 MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 29 
 
 of these studies. The oenotheraceous bud offers distinctive characters for 
 the separation of the several species (plate 8). It has, as is well known, three 
 distinct regions, the ovary, the hypanthium, and the cone — which are in most 
 cases marked off from each other in such a way as to make their accurate 
 measurement easy. As proof of this fact it is only necessary to say that fre- 
 quent repetitions 
 of the measure- 
 ments showed only 
 a variation of one- 
 tenth of a milli- 
 meter in the thick- 
 ness of ovary and 
 hypanthium, and 
 from o.2too.3mm. 
 in thickness of the 
 cone and lengths of 
 the several regions, 
 the maximum error 
 in the smallest 
 me asurements 
 taken being less 
 than 4 per cent of 
 the magnitude 
 measured, and in 
 the larger measure- 
 ments, such as 
 length of hypanthi- 
 um and cone, less 
 than 1 per cent. 
 
 In deciding to 
 make statistical 
 studies on buds a 
 careful preliminary 
 survey was made 
 to determine what 
 precautions would 
 need to be taken 
 
 in order to free the work from error. The several points that seemed to need 
 consideration were : (a) A probable periodicity in the size of the buds, so that 
 those in the early part of the season are doubtless a little larger than those 
 produced near the end of the season, but as the flowering season is very long this 
 periodicity is certainly slight and would probably have no appreciable influence 
 
 
 1 1 
 
 X„ 
 
 -4 
 
 -3 
 
 -2 -1 
 
 
 
 1 
 
 2 
 
 3 
 
 4 
 
 5 6 
 
 7 8 
 
 9 
 
 10 
 
 u 
 
 12 
 
 13 
 
 14 15 
 
 
 
 
 
 
 
 
 i5 
 
 50 
 
 55 
 
 GO 
 
 65 
 
 70 
 
 75 
 
 80 
 
 85 
 
 90 95 
 
 100|105 
 
 110 
 
 115 
 
 120 
 
 125 
 
 130 
 
 135 
 
 HO 
 
 
 
 
 
 
 
 
 50 
 
 00 
 
 60|05 
 
 70 
 
 75 
 
 80 
 
 85 
 
 00 
 
 95 100 
 
 105 
 
 110 
 
 115 
 
 120 
 
 125 
 
 130 
 
 135 
 
 140 
 
 145 
 
 
 
 
 
 Xi 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -11 
 
 16 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 -10 
 
 n 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -9 
 
 18 
 
 
 1 
 
 1 
 
 
 1 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 -8 
 
 19 
 
 
 
 1 
 
 1 
 
 1 
 
 1 
 
 1 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 6 
 
 
 
 -7 
 
 20 
 
 
 
 4 
 
 1 
 
 4 
 
 4 
 
 1 
 
 2 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 le 
 
 
 
 -G 21 
 
 |3 
 
 7 
 
 9 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 2X 
 
 
 
 -5,22 
 
 
 2 1 3 ! 13 
 
 12 
 
 7 
 
 5 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 43 
 
 
 
 -4|23 
 -3j24 
 
 
 15 [14 
 
 16 
 
 9 
 
 5 
 
 1 
 
 4 
 
 I 
 
 
 
 
 
 
 
 
 
 
 
 
 
 55 
 
 
 1 5 1 12, 12 i 13 
 
 5 
 
 
 
 4 
 
 
 2 
 
 2 
 
 
 
 
 
 
 
 
 
 
 
 59 
 
 
 
 -2125 
 
 ' ! 3 
 
 23 20 
 
 33 
 
 17 
 
 13 
 
 6 
 
 1 
 
 
 1 
 
 1 
 
 
 
 
 
 
 
 
 
 
 118 
 
 
 
 -I 
 
 26 
 
 i 1 -1 
 
 10 29 
 
 19 
 
 15 
 
 5 
 
 8 
 
 5 
 
 
 
 
 
 
 
 
 
 
 
 
 
 95 
 
 
 
 
 
 27 
 
 
 1 j 4 ! 9 '21' 31 
 
 18 
 
 14 
 
 7 
 
 G 
 
 4 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 119 
 
 
 
 1 
 
 28 
 
 
 1 ! S . 18 , 2'J 
 
 32 
 
 10 
 
 10; G 
 
 1 
 
 5 
 
 1 
 
 
 
 
 
 
 
 
 
 
 120 
 
 
 
 2 
 
 29 
 
 
 I 1 
 
 10 [16 
 
 OO 
 
 5 
 
 5 
 
 3 
 
 2 
 
 3 
 
 
 
 
 
 
 
 
 
 
 
 07 
 
 
 13 
 
 30 
 
 1 ' 1 ' 2 
 
 9 
 
 18 
 
 13 
 
 12 
 
 7 
 
 6 
 
 4 
 
 1 
 
 2 
 
 1 
 
 2 
 
 
 
 
 
 
 
 
 79 
 
 
 I4 
 
 31 
 
 
 
 1 1 
 
 4 
 
 7 
 
 III 12 
 
 11 
 
 6 
 
 3 
 
 1 
 
 1 
 
 1 
 
 
 
 
 1 
 
 
 
 
 
 59 
 
 
 
 5 
 
 32 
 
 
 
 
 3 
 
 1 
 
 G 
 
 12 
 
 S 
 
 3 
 
 4 
 
 
 
 
 3 
 
 
 
 
 
 
 
 
 
 
 39 
 
 
 
 G 
 
 33 
 
 
 I 
 
 
 1 
 
 1 
 
 7 
 
 7 
 
 13 
 
 7 
 
 3 
 
 1 
 
 1 
 
 2 
 
 
 
 
 1 
 
 
 
 
 1 
 
 
 4G 
 
 
 
 7 
 
 34 
 
 
 
 
 
 2 
 
 4 
 
 3 
 
 4 
 
 4 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 IS 
 
 
 
 8 
 
 35 
 
 
 
 
 
 2 
 
 1 
 
 3 
 
 2 
 
 1 
 
 1 
 
 3 
 
 
 
 
 
 
 
 
 
 
 
 
 13 
 
 
 
 9 
 
 36 
 
 
 
 
 
 
 
 2 
 
 
 1 
 
 2 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 6 
 
 
 
 10 
 
 .37 
 
 
 
 
 
 
 
 1 
 
 1 
 
 1 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 i 
 
 
 
 11 
 
 38 
 
 
 
 
 
 
 
 1 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 2 
 
 
 
 12' 39 
 
 
 
 
 
 
 
 
 
 
 2 
 
 
 
 
 
 
 
 
 
 
 
 
 
 2 
 
 
 
 13 40 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 h'ji 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 15 
 
 42 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 16 
 
 43 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 17 
 
 44 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 18 
 
 45 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 19 
 
 46 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 20 
 
 47 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 7 
 
 3t 
 
 lOfl 
 
 172 
 
 209il7l 
 
 1U2 
 
 S3 
 
 47 
 
 21 
 
 17 
 
 12 
 
 2 
 
 3 
 
 
 
 1 
 
 1 
 
 
 
 
 
 1 
 
 
 995 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 P'lG. 8. — Correlation between width (subject )and length of leaf in Oenolliera 
 rubrinenis. Measurements in millimeters. Coefficient of correlation, 
 0.4452±0.0171. 
 
30 MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 
 
 during the greater part of the season if the earUest and latest flowers are 
 omitted, (b) A variation in the size of the buds in response to different con- 
 ditions of the weather. While this variation is not often great enough to be 
 noticeable, there were times at which the buds seemed distinctly larger after 
 several days of warm, humid weather. Both of these difficulties would be 
 rendered comparatively harmless by a collection of a small number of buds 
 each day from each of the lots of plants studied, so that all were being subjected 
 to the same conditions. There are still some elements of error, however, 
 which this method does not fully meet, such, for instance, as the existence of a 
 different period in the different species, or a different kind or degree of reaction 
 to changes in the weather. Without an extensive study of the periodicity and 
 
 response in the several 
 species to changed cli- 
 matic conditions it can 
 not be known just how 
 serious errors these 
 factors introduce, but 
 it is believed that they 
 are slight, (c) There 
 is always present a 
 complete series of 
 buds, beginning with 
 those ready to bloom 
 and ending with those 
 just beginning to de- 
 velop, and the develop- 
 ment is so rapid that 
 buds collected at one 
 hour of the day are 
 not the equivalent of those collected at any other hour. It was found on 
 examination that the buds which would bloom the following day could be 
 readily distinguished by their size and color. Rarely there would be a doubt 
 about some particular bud, and such were omitted from the study. To 
 avoid the collection of buds at different stages of development they were 
 collected from all the lots at 5 p. m. and measured as soon thereafter as pos- 
 sible. This makes all the buds used in these studies as nearly comparable as 
 could well be secured, though it is realized that a source of slight error is to be 
 found in the fact that on some days the development of the buds is more 
 rapid than on others, so that they are not always quite equivalent at the time 
 of collection, when this is determined by the clock without regard to the 
 weather. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Xo 
 
 -11 
 
 -10 
 
 -9 
 
 -8 
 
 -7 
 
 -6 
 
 -5 
 
 -4 
 
 -3 
 
 _o 
 
 -1 
 
 
 
 1 
 
 2 
 
 3 
 
 4 
 
 5 
 
 G 
 
 7 
 
 
 
 
 
 
 
 
 o» 
 
 •23 
 
 24 
 
 ■25 
 
 26 
 
 •1- 
 
 28 
 
 29 
 
 30 
 
 31 
 
 32 
 
 33 
 
 34 
 
 3o 
 
 3i; 
 
 37 
 
 38 
 
 39 
 
 40 
 
 
 
 
 Xi 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -6 
 
 60 
 
 65 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 -5 
 
 65 
 
 70 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -4 
 
 70 
 
 75 
 
 
 1 
 
 
 
 
 
 
 1 
 
 
 1 
 
 
 1 
 
 
 1 
 
 
 
 
 
 
 
 
 5 
 
 
 -3 
 
 75 
 
 80 
 
 
 
 
 
 1 
 
 
 
 
 2 
 
 
 
 
 I 
 
 
 
 
 
 1 
 
 
 
 
 5 
 
 
 _2 
 
 80 
 
 S5 
 
 
 
 
 
 1 
 
 1 
 
 5 
 
 2 
 
 
 1 
 
 3 
 
 
 2 
 
 1 
 
 4 
 
 1 
 
 
 
 1 
 
 1 
 
 
 
 25 
 
 
 -1 
 
 85 
 
 00 
 
 
 
 
 
 
 
 1 
 
 1 
 
 1 
 
 3 
 
 1 
 
 4 
 
 1 
 
 
 
 5 
 
 
 3 
 
 
 1 
 
 
 
 23 
 
 
 
 
 90 
 
 95 
 
 
 
 
 1 
 
 
 1 
 
 
 
 
 
 1 
 
 4 
 
 2 
 
 1 
 
 5 
 
 3 
 
 4 
 
 
 
 2 
 
 
 
 
 
 30 
 
 
 1 
 
 05 
 
 IOC 
 
 
 
 
 
 
 2 
 
 1 
 
 3 
 
 
 
 
 
 5 
 
 1 
 
 
 G 
 
 4 
 
 4 
 
 5 
 
 5 
 
 1 
 
 1 
 
 
 42 
 
 
 2 
 3 
 
 100 
 
 1C5 
 
 1C6 
 110 
 
 
 1 
 
 
 
 
 
 1 
 
 
 
 
 1 
 
 4 
 
 
 3 
 
 2 
 
 
 
 3 
 
 2 
 
 
 
 17 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 2 
 
 1 
 
 
 4 
 
 1 
 
 2 
 
 
 10 
 
 
 i 
 
 110 
 
 115 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 1 
 
 
 5 
 
 115 
 
 120 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 2 
 
 1 
 
 1 
 
 
 
 4 
 
 S 
 
 9 
 
 5 
 
 8 
 
 14 
 
 12 
 
 5 
 
 IS 
 
 20 
 
 10 
 
 13 
 
 17 
 
 8 
 
 3 
 
 
 iGd 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Fig. 9. — Correlation in the buds oi OenotheralamarckianaiTomHilveT- 
 sum. Length of ovary subject, length of hypanthium relative. 
 Coefficient of correlation, 0.3872 zb 0.0453. 
 
MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 3 1 
 
 This work was not begun until late in the season, the first collection of buds 
 for this study having been made on September 1 5 and the last on October 30, 
 and this will need to be taken into account in making future comparisons, 
 though it seems likely that the periodicity is too slight to make this a modify- 
 ing factor of any considerable moment. 
 
 The number of buds from each lot of specimens, except one, was from 140 
 to 161. In Oenothera gigas only 91 buds were secured before frost cut short 
 the investigation. The lateness of the time at which the work was begun 
 prevented the measurement of a larger number of buds of each lot, though it is 
 realized that a much larger number would be desirable. On this account it 
 is not thought necessary to present variation curves of each lot of measure- 
 ments. Instead, the constants have been calculated and are presented in the 
 form of a table : 
 
 
 Length of ovary. 
 
 Length of hypanthium. 
 
 Mean. 
 
 Standard 
 
 Coefficient 
 
 Mean. 
 
 vStandard 
 
 Coefficient 
 
 
 deviation. 
 
 of variation. 
 
 deviation. 
 
 of variation. 
 
 Deci-millimeters. 
 
 Deci- m illimete rs 
 
 Per cent. 
 
 Milliineters. 
 
 Millimeters. 
 
 Per cent. 
 
 a 
 
 95 67 ±0 50 
 
 9-33 ±0.35 
 
 9-75±o.39 
 
 3306 ±0.23 
 
 4.20 ±0.16 
 
 I2.7I±0.49 
 
 b 
 
 91 76 ±0.47 
 
 S.73 ±0.33 
 
 9.52±o.36 
 
 33 07 ±0.22 
 
 4 
 
 18 ±0.16 
 
 i2.63±o.48 
 
 c'* 
 
 92.29 ±0 34 
 
 6 40 ±0.24 
 
 6 94±o.26 
 
 29.10 ±0.22 
 
 4 
 
 14 ±0.16 
 
 i4.23±o.55 
 
 c 
 
 10756 ±0.53 
 
 9.89 ±0.37 
 
 9-20±o.35 
 
 30.37 ±0.21 
 
 3 
 
 88 ±0.15 
 
 i2.77±o.49 
 
 d 
 
 Ti8.8i ±0.47 
 
 8.40 ±0.33 
 
 707±o. 28 
 
 30.33 ±0.18 
 
 3 
 
 24 ±0.13 
 
 io.69±o.43 
 
 e 
 
 102.12 ±0.64 
 
 9.01 ±0.-45 
 
 8.83±o.44 
 
 25-18 ±0.37 
 
 5 
 
 29 ±0.26 
 
 2 1 . 00 ± 1 . 09 
 
 f 
 
 83.42 ±0.44 
 
 7.77 ±0.31 
 
 9-3i±o.38 
 
 26.96 ±0.32 
 
 5 
 
 61 ±0.23 
 
 2o.8o±o.87 
 
 
 Length of cone. 
 
 Thickness of ovary. 
 
 Millimeters. 
 
 Millimeters. 
 
 
 Deci-millimeters. 
 
 Deci-millimeters 
 
 
 a 
 
 39-47 ±0.19 
 
 3-47 ±0.13 
 
 8.79±o.34 
 
 32-37 ±0.10 
 
 I. 932 ±0.073 
 
 5-97±o.23| 
 
 b 
 
 39 
 
 47 ±0.18 
 
 ^ 
 
 45 ±0.13 
 
 8.74±o.33 
 
 33.51 ±0.11 
 
 2.o78±o.078 
 
 6 
 
 20±0.24 
 
 c' 
 
 37 
 
 80 ±0.14 
 
 2 
 
 65 ±0.10 
 
 7 .01 ±0.26 
 
 31 .261 ±0.091 
 
 1 .711 ±0.064 
 
 5 
 
 47±o.2i 
 
 c 
 
 35 
 
 52 ±0 17 
 
 3 
 
 29 ±0.15 
 
 9.26±o.35 
 
 33-76 ±0.14 
 
 2 .540 ±0.096 
 
 7 
 
 52±0.28 
 
 d 
 
 36 
 
 35 ±0.13 
 
 2 
 
 277 ±0.090 
 
 6.26±o.25 
 
 36.48 ±O.TI 
 
 2.041 ±0.081 
 
 5 
 
 59 ±0 22 
 
 e 
 
 39 
 
 51 ±0.36 
 
 5 
 
 29 ±0.26 
 
 i2.75±o.65 
 
 44.46 ±0.31 
 
 3.69 ±0.22 
 
 8 
 
 3o±o.49 
 
 t 
 
 30 
 
 27 ±0.24 
 
 4 
 
 27 ±0.17 
 
 i409±o.58 
 
 38.24 ±0.11 
 
 i.938±o.o78 
 
 5.o7±o 20 
 
 — 
 
 Thickness of hypanthium. 
 
 Thickness of cone. 
 
 Deci-millimeters. \ Deci-millimeters 
 
 
 Deci-millimeters. 
 
 Deci-millimeters 
 
 
 a 
 
 25 291 ±0.068 r .261 ±0.048 
 
 4-99±o 19 
 
 84.59 ±0 39 
 
 7.21 ±0.27 
 
 8.53±o.33 
 
 b 
 
 26.o69±o.o7o 
 
 I 
 
 .32o±o.o5o 
 
 5.o6±o. 19 
 
 81 
 
 81 ±0.32 
 
 6.00 ±0.23 
 
 7 
 
 33±o.28 
 
 c' 
 
 25-453±o.o74 
 
 I 
 
 389±o.o52 
 
 5-46±o.2i 
 
 87 
 
 60 ±0.35 
 
 6.63 ±0.25 
 
 / 
 
 56±o.29 
 
 c 
 
 28.46 ±0. 10 
 
 I 
 
 940 ±0.073 
 
 6.82±o.26 
 
 93 
 
 26 ±0 43 
 
 8.01 ±0.30 
 
 8 
 
 59±o.33 
 
 d 
 
 30-356±o.o70 
 
 I 
 
 259 ±0 050 
 
 4- i5±o.i6 
 
 93 
 
 47 ±0.29 
 
 5.17 ±0.20 
 
 5 
 
 53±o.2i 
 
 e 
 
 36.45 ±0.24 
 
 3 
 
 46 ±0.17 
 
 9. 48 ±0.48 109 
 
 76 ±0.69 
 
 9-79 ±0.49 
 
 8 
 
 92±o.45 
 
 f 
 
 29- i57±o.o8i 
 
 I 
 
 422±o.o57 
 
 4.88±o 20 
 
 96 
 
 41 ±0.47 
 
 8.18 ±0.33 
 
 8.48±o.34 
 
 *cf ^Oenothera lamarckiana from lot c. 
 
32 MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 
 
 One of the things of first interest that will be looked for in this table is the 
 relation of the fluctuation found in different lots of plants of the same species. 
 Particularly is it interesting to compare the conditions of entry a with entry 
 b, the first representing Oenothera lamarckiana self-fertilized for sixteen genera- 
 tions, the second the same species which has never been self-fertilized, except 
 as that process may have taken place by natural causes. 
 
 One represents i6 generations of artificial selection, the other many genera- 
 tions of natural selection, but in neither case is it probable that the propor- 
 tionality in the parts of the bud entered into the selection in any specific way. 
 In the mean values of the several parts, these two lots of plants are found to be 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Xo 
 
 -13 
 
 -VI 
 
 Ui 
 
 -10 
 
 -9 
 
 -8 
 
 ~ i 
 
 -G 
 
 -5 
 
 -4 
 
 -3 
 
 — o 
 
 -1 
 
 
 
 1 
 
 2 
 
 3 
 
 4 
 
 5 
 
 G 
 
 7 
 
 8 
 
 9 
 
 10 
 
 
 
 
 
 
 
 
 20 
 
 21 
 
 22 
 
 23 
 
 24 
 
 25 
 
 20 
 
 27 
 
 28 
 
 29 
 
 30 
 
 31 
 
 32 
 
 33 
 
 34 
 
 35 
 
 3G 
 
 37 
 
 38 
 
 39 
 
 40 
 
 41 
 
 42 
 
 43 
 
 
 
 
 X, 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -11 
 
 35 
 
 40 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 -10 
 
 to 
 
 45 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -y 
 
 45 
 
 50 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -s 
 
 50 
 
 55 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ~< 
 
 55 
 
 60 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -fi 
 
 60 
 
 65 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -5 
 
 65 
 
 70 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -4 
 
 70 
 
 75 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -3 
 
 75 
 
 SO 
 
 
 1 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 2 
 
 
 
 
 
 
 
 
 
 
 
 
 
 4 
 
 
 _o 
 
 SO 
 
 85 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 2 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 3 
 
 
 -1 
 
 
 
 85 
 90 
 
 90 
 
 95 
 
 
 
 
 
 
 
 2 
 
 1 
 
 1 
 
 o 
 
 2 
 
 2 
 
 1 
 
 3 
 
 5 
 
 
 2 
 
 
 1 
 
 
 
 
 
 
 
 
 •n 
 
 
 
 
 
 1 
 
 
 
 1 
 
 1 
 
 O 
 
 2 
 
 4 
 
 1 
 
 3 
 
 4 
 
 G 
 
 
 4 
 
 o 
 
 5 
 
 1 
 
 1 
 
 
 
 
 
 
 42 
 
 
 1 
 
 95 
 
 lUO 
 
 
 
 
 
 
 1 
 
 
 2 
 
 1 
 
 
 3 
 
 1 
 
 2 
 
 3 
 
 4 
 
 
 2 
 
 6 
 
 i 
 
 2 
 
 1 
 
 1 
 
 1 
 
 
 
 
 3S 
 
 
 o 
 
 100^05 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 1 
 
 2 
 
 3 
 
 1 
 
 
 2 
 
 o 
 
 o 
 
 2 
 
 2 
 
 
 
 
 
 
 19 
 
 
 3 
 
 lOO'llO 
 
 
 
 
 
 
 
 
 
 
 
 
 2 
 
 1 
 
 2 
 
 4 
 
 
 1 
 
 
 
 1 
 
 
 3 
 
 1 
 
 
 1 
 
 
 17 
 
 
 4 
 
 110 
 
 115 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 1 
 
 2 
 
 
 1 
 
 2 
 
 
 
 
 
 
 7 
 
 
 5 
 
 115 
 
 120 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 1 
 
 
 
 1 
 
 
 
 
 
 
 3 
 
 
 G 
 
 120 
 
 125 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 I 
 
 
 
 1 
 
 
 
 O 
 
 4 
 
 4 
 
 4 
 
 4 
 
 13 
 
 7 
 
 11 
 
 17 
 
 20 
 
 8 
 
 12 
 
 13 
 
 15 
 
 7 
 
 7 
 
 4 
 
 3 
 
 
 
 1 
 
 158 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Fig. 10. — Correlation in the buds of OenoUiera lamarckiana, Xo. 04129. Length of 
 ovary subject, length of hypanthium relative. Coefficient of correlation, 0.4215 ± 
 
 0.0441. 
 
 identical only in regard to length of hypanthium and length of cone. Consid- 
 erable differences are to be seen in length of ovary and thickness of cone, and 
 slight differences in thickness of ovary and of hypanthium. In the standard 
 deviation and coefficient of variability, on the other hand, there is essential 
 identity in every character except in the thickness of the cone, which was 
 significantly more variable in the self-fertilized plants. There are too many 
 possible sources of such difference to make any speculation regarding its 
 significance profitable at this time. 
 
 When the data in the third entry (c'), is compared with those in the first 
 and fourth entries, the very interesting fact develops that in certain charac- 
 
MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE) OENOTHERAS. 33 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Xo 
 
 -9 
 
 -S 
 
 ~ J 
 
 -G 
 
 -5-4 
 
 -3 
 
 —2 
 
 -1 
 
 
 
 1 
 
 
 
 3 
 
 4 
 
 5 
 
 
 
 7 
 
 S 
 
 9 
 
 lu 
 
 U 
 
 
 
 
 
 
 
 
 ■20 
 
 •71 
 
 00 
 
 ■23 
 
 24 '25 
 
 ■2G 
 
 27 
 
 2S 
 
 29 
 
 30 
 
 31 
 
 32 
 
 33 
 
 34 
 
 35 
 
 3G 
 
 37 
 
 3S 
 
 39 
 
 40 
 
 
 
 
 X, 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -4 
 
 70 
 
 75 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 "■J 
 
 75 
 
 80 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 —2 
 
 SO 
 
 85 
 
 
 1 
 
 
 
 
 2 
 
 1 
 
 
 2 
 
 2 
 
 1 
 
 1 
 
 
 
 I 
 
 
 2 
 
 
 
 
 
 
 
 
 15 
 
 
 -1 
 
 Ho 
 
 90 
 
 
 I 
 
 
 1 
 
 3 
 
 / 
 
 G 
 
 5 
 
 3 
 
 4 
 
 3 
 
 i 
 
 1 
 
 5 
 
 1 
 
 3 
 
 
 
 
 
 
 
 
 47 
 
 
 
 
 uo 
 
 <J5 
 
 
 
 1 
 
 1 
 
 
 1 
 
 2 
 
 4 
 
 - 
 
 4 
 
 1 
 
 1 
 
 3 
 
 3 
 
 
 
 3 
 
 1 
 
 
 
 1 
 
 
 
 
 
 40 
 
 
 1 
 
 95 
 
 100 
 
 
 
 
 
 
 3 
 
 
 3 
 
 2 
 
 4 
 
 •J 
 
 5 
 
 5 
 
 
 
 2 
 
 '> 
 
 5 
 
 
 
 
 
 
 
 37 
 
 
 
 
 100 
 
 105 
 
 
 
 
 
 
 
 
 
 1 
 
 1 
 
 
 2 
 
 1 
 
 1 
 
 1 
 
 1 
 
 1 
 
 
 1 
 
 1 
 
 1 
 
 
 
 12 
 
 
 3 
 
 105 
 
 IIU 
 
 
 
 
 
 
 
 
 1 
 
 
 1 
 
 
 
 
 1 
 
 2 
 
 
 2 
 
 
 1 
 
 
 
 1 
 
 
 9 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 3 
 
 2 
 
 5 
 
 1-2 
 
 8 
 
 15 
 
 16 
 
 15 
 
 8 
 
 15 
 
 lU 
 
 14 
 
 8 
 
 11 
 
 9 
 
 2 
 
 3 
 
 1 
 
 1 
 
 1 
 
 
 ^61 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Fig. 11 .—Correlation in the buds of Oenothera lamarckiana. No. 04113. 
 Length of ovary subject, length of hypanthium relative. Coefficient of 
 correlation, 0.4719 ± 0.041.?. 
 
 ters, as thickness of hypanthium and length and thickness of cone, the plants 
 here represented stand intermediate between those of a and c, and that the 
 mean length of hypanthium was even less than in either, though it should 
 have been expected 
 to be as great as 
 that of a ; in other 
 words, the speci- 
 mens oiO. lamarck- 
 iana which sprang 
 from a rubrinervis 
 parent had certain 
 characters interme- 
 diate between pure- 
 bred 0. lamarckiana 
 and O. rubrinervis, 
 and in length of 
 hypanthium they 
 were even less than 
 
 those of the latter species. These plants were considered typical Oenothera 
 lamarckiana, not only by myself, but by others who are familiar with that 
 species. Here again, as in the rest of the data presented in this table, the 
 general validity of any conclusion reached may be questioned because of 
 the small amount 
 of data. 
 
 Comparison of 
 the entries c and 
 d, representing the 
 variation in the 
 buds of two dif- 
 ferent pedigrees of 
 0. rubrinervis, 
 shows that only in 
 length of hypan- 
 thium and thick- 
 ness of cone were 
 the two lots 
 essentially alike 
 in mean value, 
 the means of the 
 other parts differ- 
 ing from each other by more than the sums of the probable errors. In respect 
 to variability, entry c shows considerably greater coefficients in every set of 
 
 
 
 
 
 
 
 
 
 
 
 ~ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Xo 
 
 rll 
 
 -10 
 
 -9 
 
 -8 
 
 —7 
 
 -6 
 
 -5 
 
 -•4 
 
 -3 
 
 _ 
 
 -1 
 
 
 
 1 
 
 2 
 
 3 
 
 4 
 
 5 
 
 G 
 
 7 
 
 8 
 
 9 
 
 10 
 
 
 
 
 
 
 
 
 •2is 
 
 •26 
 
 30 
 
 31 
 
 32 
 
 33 
 
 34 
 
 35 
 
 36 
 
 37 
 
 38 
 
 39 
 
 40 
 
 41 
 
 42 
 
 43 
 
 44 
 
 45 
 
 4G 
 
 47 
 
 48 
 
 49 
 
 
 
 
 Xi 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -6 
 
 60 
 
 C5 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 -5 
 
 05 
 
 70 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -4 
 
 Hi 
 
 75 
 
 
 
 
 
 
 I 
 
 1 
 
 1 
 
 
 
 1 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 5 
 
 
 -3 
 
 75 
 
 80 
 
 
 
 
 
 
 
 
 2 
 
 
 2 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 5 
 
 
 _o 
 
 80 
 
 85 
 
 
 
 1 
 
 
 
 
 
 
 2 
 
 4 
 
 3 
 
 3 
 
 1 
 
 3 
 
 4 
 
 2 
 
 
 
 
 
 
 
 
 
 
 25 
 
 
 -1 
 
 85 
 
 90 
 
 
 
 
 
 
 
 
 
 
 2 
 
 2 
 
 2 
 
 G 
 
 2 
 
 2 
 
 2 
 
 
 1 
 
 1 
 
 1 
 
 
 
 
 
 
 OC 
 
 
 
 
 90 
 
 95 
 
 
 
 
 
 
 
 
 1 
 
 1 
 
 5 
 
 1 
 
 3 
 
 •■» 
 
 4 
 
 5 
 
 5 
 
 -' 
 
 1 
 
 
 
 
 
 
 
 3J 
 
 
 1 
 
 95 
 
 100 
 
 
 
 
 
 
 
 
 1 
 
 2 
 
 '> 
 
 4 
 
 4 
 
 3 
 
 7 
 
 <i 
 
 5 
 
 5 
 
 5 
 
 
 1 
 
 
 
 
 
 42 
 
 
 3 
 
 IOC 
 
 1C5 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 2 
 
 2 
 
 3 
 
 
 
 
 1 
 
 3 
 
 1 
 
 
 
 1 
 
 1 
 
 
 17 
 
 
 1C5 
 
 no 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 1 
 
 1 
 
 5 
 
 1 
 
 
 
 1 
 
 
 10 
 
 
 4 
 
 no 
 
 115 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 1 
 
 
 5 
 
 115 
 
 ICO 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 1 
 
 
 
 
 
 1 
 
 1 
 
 7 
 
 7 
 
 IG 
 
 11 
 
 19 
 
 10 
 
 20 
 
 IG 
 
 13 
 
 12 
 
 11 
 
 7 
 
 2 
 
 
 
 1 
 
 
 
 
 lUO 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Fig. 12. — Correlation in the buds of Oenothera lamarckiana from Hilversum. 
 Length of ovary subject, length of bud-cone relative. Coefficient of correla- 
 tion, 0.6340± 0.0319. 
 
34 MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 
 
 measurements than entry d. The two lots of plants here compared may be 
 contrasted with respect to their origin and treatment as follows, and seme 
 one, several, or all of these contrasted factors will probably satisfactorily 
 account for the difference in variability: 
 
 c. 
 
 d. 
 
 Seed from the New York Botanical Gar- 
 den. 
 
 Of hybrid (?) origin, 0. lamarckiana X 
 biennis. 
 
 PolHnation not guarded. 
 
 Seeds sown Feb. i6, 1905. 
 
 Buds collected Sept. 15 to 28, being the 
 last half of the season. 
 
 Seed from the Botanical Garden of Am- 
 sterdam. 
 From pure-bred 0. lamarckiana. 
 
 Pollination guarded, self-fertilized. 
 Seeds sown Mar. 11, 1905. 
 Buds collected Sept. 30 to Oct. 10, being 
 early in the flowering season. 
 
 It remains to compare the three mutants here represented — 0. ruhrinervis, 
 0. gigas, and 0. lata — ^with the parent species in regard to mean values of 
 their several characters and the coefficients of their variability. This can 
 
 not be done with entire 
 satisfaction because of the 
 inconsistencies in the meas- 
 urements of plants belong- 
 ing to different pedigrees 
 in the same species. Such 
 comparison would be just 
 only when the lots of mate- 
 rial of all the species had 
 received identical treat- 
 ment, and this, as we have 
 seen, has not been true of 
 these. Wherever several 
 lots of plants of a single 
 species have been used, the 
 average condition of the 
 several lots has been taken 
 for comparison, and as a 
 given amount of deviation 
 in a constant of one value 
 is unfairly compared with 
 a given deviation from a 
 constant of a different value, all the deviations have been reduced to per cents 
 of the equivalent measurement in Oenothera lamarckiana. These values are 
 arranged in the following table in such a way as to allow a comparison 
 
 
 
 
 
 
 
 
 
 ~ 
 
 
 
 
 
 r 
 
 
 
 
 
 
 
 
 
 
 
 
 
 x. 
 
 -7 
 
 -(1 
 
 -5 
 
 -1 
 
 -3 
 
 -2 
 
 -1] 
 
 1 
 
 
 
 :; 
 
 4 
 
 .5 
 
 6 
 
 7 
 
 8 
 
 9 
 
 
 
 
 
 
 
 
 32 
 
 33 
 
 31 
 
 35 
 
 30 
 
 37 
 
 38 
 
 3'J 
 
 W 
 
 11 
 
 12 
 
 13 
 
 44 
 
 45 
 
 46 
 
 47 
 
 48 
 
 
 
 
 Xi 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -11 
 
 3^ 
 
 to 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 -10 
 
 10 
 
 15 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -9 
 
 Ij 
 
 50 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -h 
 
 5(J 
 
 55 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -7 
 
 o5 
 
 60 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -6 
 
 60 
 
 65 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -5 
 
 65 
 
 70 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 — ! 
 
 70 
 
 75 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -3 
 
 75 
 
 80 
 
 
 1 
 
 1 
 
 
 
 1 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 4 
 
 
 
 
 80 
 
 85 
 
 
 
 
 1 
 
 
 
 
 1 
 
 
 I 
 
 
 
 
 
 
 
 
 
 
 3 
 
 
 -1 
 
 85 
 
 90 
 
 
 2 
 
 1 
 
 
 3 
 
 3 
 
 3 
 
 5 
 
 3 
 
 
 1 
 
 2 
 
 
 
 
 
 
 
 
 23 
 
 
 
 
 91, 
 
 95 
 
 
 1 
 
 
 
 -1 
 
 1 
 
 10 
 
 6 
 
 9 
 
 A 
 
 2 
 
 2 
 
 2 
 
 1 
 
 
 
 
 
 
 42 
 
 
 1 
 
 95 
 
 100 
 
 
 
 I 
 
 1 
 
 1 
 
 3 
 
 5 
 
 5 
 
 5 
 
 7 
 
 3 
 
 4 
 
 
 
 
 
 1 
 
 
 
 
 38 
 
 
 .-) 
 
 ICO 
 
 105 
 
 
 
 
 
 
 1 
 
 1 
 
 
 
 
 
 
 
 
 
 4 
 
 1 
 
 
 1 
 
 1 
 
 
 
 19 
 
 
 3 
 
 105 
 
 110 
 
 
 
 
 
 
 
 
 
 
 1 
 
 1 
 
 1 
 
 
 
 3 
 
 4 
 
 1 
 
 1 
 
 
 
 17 
 
 
 4 
 
 110 
 
 115 
 
 
 
 
 
 
 
 
 1 
 
 
 1 
 
 1 
 
 
 
 
 
 
 1 
 
 2 
 
 
 7 
 
 
 5 
 
 115 
 
 120 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 1 
 
 
 
 
 3 
 
 
 6 
 
 120 
 
 125 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 — 
 
 
 
 
 1 
 
 3 
 
 2 
 
 8 
 
 9 
 
 ■22 
 
 19 
 
 20 
 
 16 
 
 11 
 
 17 
 
 
 
 5 
 
 6 
 
 4 
 
 3 
 
 3 
 
 
 158 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 _J 
 
 
 
 
 
 
 
 
 
 
 Fig. 13. — Correlation in the buds of Oenothera lamarckiana. No. 
 04129. Length of ovary subject, length of bud-cone relative. 
 Coefficient of correlation, 0.5631 ± 0.0366. 
 
MUTATIONS, VARIATIONS, AND REIyATIONSHIPS OF THE OENOTHERAS. 35 
 
 between the deviation in the mean, dM, and that of the coefficient of vari- 
 abihty, dV , not only of the characters studied in 1905, but also those of 
 the preceding year derived in the same way. 
 
 Deviation of means and variabilities, in per cents of the corresponding values for Oenothera 
 
 lamarckiana. 
 
 1904. 
 
 O. rubrinervis 
 Length of leaf . , 
 
 Width of leaf 
 
 I-eaf ratio 
 
 Number of branches. 
 Total branch length . 
 
 O. nanella, 1904. 
 Height 
 
 O. rubrinervis, igo$. 
 
 Length of leaf 
 
 Width of leaf 
 
 Leaf ratio 
 
 Height 
 
 Number of branches. . . 
 Total branch length . . 
 
 Length of ovary 
 
 Length of hypanthium. 
 Length of cone 
 
 dM 
 
 dV 
 
 1 .02 
 
 - 7.21 
 
 20 . 95 
 
 6.91 
 
 24-75 
 
 8.08 
 
 104-59 
 
 -4.46 
 
 172.30 
 
 116.34 
 
 74.28 
 
 492 . 90 
 
 20. 19 
 
 — 22.26 
 
 25-33 
 
 2.89 
 
 6.93 
 
 16.95 
 
 5-98 
 
 11.23 
 
 15-19 
 
 33-22 
 
 4-57 
 
 36.72 
 
 21.38 
 
 - 6.99 
 
 4-38 
 
 — 11.07 
 
 7.66 
 
 - 5.13 
 
 0. rubrinervis, igo5. 
 
 Thickness of ovary 
 
 Thickness of hypanthium 
 Thickness of cone 
 
 O. gigas, 1905. 
 
 Length of ovary 
 
 Length of hypanthium. . . 
 
 Length of cone ... 
 
 Thickness of ovary 
 
 Thickness of hypanthium 
 Thickness of cone 
 
 O. lata, 1905. 
 
 Length of ovary . 
 
 Length of hypanthium. . . 
 
 Length of cone 
 
 Thickness of ovary 
 
 Thickness of hypanthium 
 Thickness of cone 
 
 dM 
 
 dV 
 
 8 46 
 
 II .40 
 
 14.84 
 
 6.00 
 
 10.26 
 
 — 9.60 
 
 9 52 
 
 1.03 
 
 20.67 
 
 57-70 
 
 1.54 
 
 55-87 
 
 37-03 
 
 41 . 16 
 
 42.38 
 
 83-37 
 
 29.63 
 
 14.21 
 
 10.53 
 
 6.53 
 
 15.06 
 
 57-70 
 
 22.21 
 
 72 25 
 
 18.13 
 
 -^3-77 
 
 13-87 
 
 - 5.61 
 
 13-87 
 
 8.58 
 
 In this table all the deviations which are less than the sum of the probable 
 errors are printed in bold-face figures, and may be left out of the discussion 
 as having no significance. If the signs of the remaining numbers are taken 
 into account it is found that 16 of the 30 characters studied showed a greater 
 variability in the mutant than in the parent, while only 4 were significantly 
 less variable in the mutant. Thus it is found again, as in 1904, that there is a 
 well-marked tendency for the mutant to show a greater variability than its 
 parent, though this relation is far from being absolute. The hint then recog- 
 nized, that the degree of departure from the mean of the parent might be 
 related to the degree of increase in the variability, is also faintly repeated, but 
 is too slight to be now considered of any significance. 
 
 The single study in the correlation of characters in 1904 showed a consid- 
 erably less correlation in the mutant than in the parent species, and we have 
 already seen in the study of leaf characters above that, though that study 
 was unsatisfactory, such evidence as there is seems to accord with the view 
 then expressed that this represents a general fact. In order to gain further 
 evidence on this point the coefficients of correlation of the parts of the buds 
 were calculated, and all the original measurements are given in the correlation 
 tables appended. 
 
36 MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE) OENOTHERAS. 
 
 The coefficients of correlation are as follows : 
 
 Length of 
 Length of 
 Length of 
 Length of 
 Length of 
 Thickness 
 Length of 
 Length of 
 Length of 
 Thickness 
 Length of 
 Thickness 
 Thickness 
 Thickness 
 Thickness 
 Thickness 
 Thickness 
 Thickness 
 
 ovary 
 
 hypanthium. 
 
 ovary 
 
 cone 
 
 ovary 
 
 of ovary. . . 
 hypanthium. . . 
 
 cone 
 
 hypanthium . . 
 of hypanthium 
 
 cone 
 
 of cone 
 
 of ovary 
 
 of hypanthium 
 
 of ovary 
 
 of cone 
 
 of hypanthium 
 of cone 
 
 a 
 
 !~ O 
 
 - o 
 
 ) 
 
 ) 
 
 r o 
 
 r o 
 
 42i±o. 
 
 563 ±o. 
 
 477±o 
 
 564±o. 
 
 i57±o, 
 
 873±o, 
 
 558±o 
 
 65i±o 
 
 5i8±o 
 
 044,0. 
 0370. 
 052 o. 
 0370. 
 052 o. 
 0130. 
 037,0. 
 031 o. 
 0390. 
 
 387±o 
 634±o 
 600 ±0 
 580 ±0 
 044 ±0 
 785±o 
 573±o 
 569 ±0 
 532±o 
 
 .0450 
 .032 o 
 
 .0340 
 ■0350 
 • 0530 
 
 .020 o 
 .036 o 
 
 .0360 
 
 I 
 
 .038;o 
 
 472±o 
 572±o 
 
 272±0, 
 
 535±o, 
 
 i33±o. 
 
 829±o, 
 
 556±o. 
 
 59i±o 
 
 500 ±0 
 
 041 o 
 
 i 
 
 0360 
 
 049 o 
 
 I 
 0380 
 
 052 o 
 
 i 
 
 017 o 
 
 1 
 
 i 
 
 o37iO 
 0350 
 040 o 
 
 298^0.049 
 598±o.o34 
 .596±o.o34 
 375±o.046 
 oi9±o.o5i 
 835±o.oi6 
 703±o.o27 
 631 ±0.032 
 ,7i7±o.026 
 
 Length of 
 Length of 
 Length of 
 Length of 
 Length of 
 Thickness 
 Length of 
 Length of 
 Length of 
 Thickness 
 Length of 
 Thickness 
 Thickness 
 Thickness 
 Thickness 
 Thickness 
 Thickness 
 Thickness 
 
 ovary 
 
 hypanthium . 
 
 ovary . 
 cone . . 
 
 / 
 
 ovary 
 
 of ovary 
 
 hypanthium. . . 
 
 cone 
 
 hypanthium . . . 
 of hypanthium . 
 
 cone 
 
 of cone 
 
 of ovary 
 
 of hypanthium . 
 
 of ovary 
 
 of cone 
 
 of hypanthium . 
 of cone 
 
 0.22I ±0. 
 0.246±0. 
 
 o. 336 ±0. 
 o.507±o. 
 o.i53±o. 
 o . 840 ± o . 
 o. 5i4±o. 
 o.5oi±o. 
 o.540±o. 
 
 0530 
 052 o 
 0500 
 
 041 o 
 
 I 
 
 055I0 
 o 
 
 016 
 042 
 042 
 
 035 
 
 094±o 
 
 ■534±o 
 .636±o 
 527±o 
 .438±o 
 456±o 
 83S±o 
 475 ±0 
 482+0 
 
 0560 
 .051 o 
 .042 o 
 .051 o 
 
 • 0570 
 •0560 
 
 . 021 o 
 
 .0750 
 .054,0 
 
 .624 ±0.043 
 .638±o.o34 
 .357. -to. 050 
 .o25±o.o57 
 .111 ±0.050 
 837±o.oi7 
 .64o±o.o34 
 .584±o.o38 
 .555±o.039 
 
 A comparison of the coefficients of correlation in the characters of the sev- 
 eral mutants with those of the corresponding characters of Oenothera lama)xk- 
 iana shows that in 14 cases out of 27 there is a decrease of correlation in the 
 mutant ; in 1 2 instances there is an increase, and in i the correlation remained 
 the same. These results seem to negative the view that the characters of the 
 Oenothera mutants are generally less closely correlated than those of the parent. 
 However, the differences between the coefficients of the mutants and those 
 of Oenothera lamarckiana are not generally greater than those between the 
 several lots of O. lamarckiana, or between the two lots of 0. riibrinervis, and 
 it is evident that the question is still to be considered undecided, but with 
 the burden of evidence in favor of the view that little or no change takes place 
 in the correlation of parts when mutation changes their proportions. 
 
MUTATIONS, VARIATIONS, AND REIvATlONSHIPS OF THE OENOTHERAS. 37 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ~" 
 
 ~ 
 
 
 
 
 
 
 
 
 Xo 
 
 ~7 
 
 -6 
 
 -5 
 
 -4 
 
 -3 
 
 — 
 
 -1 
 
 
 
 1 
 
 9 
 
 3 
 
 4 
 
 5 
 
 G 
 
 7 
 
 
 
 
 
 
 
 
 30 
 
 31 
 
 32 
 
 33 
 
 34 
 
 35 
 
 3U 
 
 37 
 
 3S 
 
 39 
 
 40 
 
 41 
 
 42 
 
 43 
 
 44 
 
 
 
 
 X, 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -4 
 
 70 
 
 75 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 -3 
 
 75 
 
 80 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ~2 
 
 80 
 
 85 
 
 
 
 
 
 1 
 
 
 5 
 
 I 
 
 4 
 
 1 
 
 
 3 
 
 
 
 
 
 
 15 
 
 
 -1 
 
 85 
 
 90 
 
 
 1 
 
 
 
 1 
 
 G 
 
 9 
 
 11 
 
 5 
 
 G 
 
 5 
 
 1 
 
 1 
 
 
 1 
 
 
 
 47 
 
 
 
 
 90 
 
 95 
 
 
 
 
 
 1 
 
 3 
 
 G 
 
 
 
 t 
 
 8 
 
 c 
 
 5 
 
 
 
 1 
 
 1 
 
 
 4u 
 
 
 1 
 
 95 
 
 100 
 
 
 
 
 
 
 
 
 
 3 
 
 5 
 
 8 
 
 8 
 
 5 
 
 
 4 
 
 Jl 
 
 
 
 37 
 12 
 
 
 2 
 
 100 
 
 105 
 
 
 
 
 
 
 
 
 
 
 1 
 
 2 
 
 ] 
 
 3 
 
 1 
 
 4 
 
 
 
 3 
 
 105 
 
 110 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 2 
 
 2 
 
 
 2 
 
 2 
 
 
 9 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 3 
 
 10 
 
 23 
 
 17 
 
 11 
 
 24 
 
 ■21 
 
 17 
 
 
 
 5 
 
 9 
 
 3 
 
 
 IGl 
 
 
 
 __ 
 
 
 
 
 _ 
 
 _ 
 
 _ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Fig. 14. — Correlation in the buds of Oenothera lamarckiana , 
 Xo. 04113. Length of ovary subject, length of bud- 
 cone relative. _CoefScient of correlation, 0.5720 ± 0.0358. 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Xo 
 
 -71-U 
 
 -5-4-:; 
 
 __o 
 
 -1 
 
 1 
 
 2 
 
 3 
 
 4 
 
 5 
 
 6 
 
 7 
 
 
 
 
 
 
 
 
 2G 27 , 2.S ] 29 : TO 
 
 31 
 
 32 
 
 33 34 
 
 35 
 
 3-3 
 
 37 
 
 38 
 
 39 
 
 40 
 
 
 
 
 Xi 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -0 
 
 60 
 
 U5 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 -5 
 
 C5 
 
 70 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -4 
 
 70 
 
 75 
 
 
 
 1 
 
 1 
 
 1 
 
 
 1 
 
 1 
 
 
 
 
 
 
 
 
 
 
 5 
 
 
 -3 
 
 75 
 
 SO 
 
 
 
 
 
 
 2 
 
 
 2 
 
 
 1 
 
 
 
 
 
 
 
 
 5 
 
 
 -2 
 
 80 
 
 86 
 
 
 
 
 
 
 4 
 
 4 
 
 5 
 
 4 
 
 4 
 
 3 
 
 1 
 
 
 
 
 
 
 25 
 
 
 -1 
 
 55 
 
 91' 
 
 
 
 
 
 
 1 
 
 
 11 
 
 1 
 
 2 
 
 4 
 
 3 
 
 
 1 
 
 
 
 
 23 
 
 
 
 
 90 
 
 95 
 
 
 
 
 
 
 
 
 
 9 
 
 5 
 
 9 
 
 2 
 
 
 
 1 
 
 
 
 
 
 30 
 
 
 1 
 
 96 
 
 ICO 
 
 
 
 
 
 
 
 
 7 
 
 9 
 
 10 
 
 9 
 
 4 
 
 3 
 
 
 
 
 
 42 
 
 
 2 
 
 100 
 
 1C6 
 
 
 
 
 
 
 
 
 1 
 
 3 
 
 3 
 
 3 
 
 4 
 
 3 
 
 
 
 
 
 17 
 
 
 3 
 
 1C5 
 
 no 
 
 
 
 
 
 
 
 
 
 2 
 
 3 
 
 2 
 
 1 
 
 1 
 
 
 
 1 
 
 
 10 
 
 
 4 
 
 110 
 
 115 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 1 
 
 
 5 
 
 115 
 
 120 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 1 
 
 1 
 
 1 
 
 7 
 
 7 
 
 36 
 
 24 
 
 32 
 
 23 
 
 15 
 
 10 
 
 1 
 
 
 
 1 
 
 
 160 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Fig. 15. — Correlation in the buds of Oenothera lamarckiana 
 from Hilversum. Length of ovary subject, thickness of 
 ovary relative. Coefficient of correlation, 0.5997 i 0.0341. 
 
38 MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 x„ 
 
 -5 
 
 -4 
 
 -3 
 
 -2 
 
 -1 
 
 
 
 1 
 
 2 
 
 3 
 
 4 
 
 5 
 
 G 
 
 
 
 
 
 
 
 
 27 
 
 28 
 
 29 
 
 30 
 
 31 
 
 32 
 
 33 
 
 34 
 
 35 
 
 3G 
 
 37 
 
 38 
 
 
 
 
 Xi 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -11 
 
 35 
 
 10 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 1 
 
 
 -10 
 
 10 
 
 15 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -9 
 
 15 
 
 50 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -8 
 
 50 
 
 55 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 — 7 
 
 55 
 
 60 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -6 
 
 60 
 
 65 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 — 
 
 65 
 
 70 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -4 
 
 70 
 
 75 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -3 
 
 75 
 
 80 
 
 
 1 
 
 1 
 
 
 1 
 
 
 1 
 
 
 
 
 
 
 
 
 4 
 
 
 O 
 
 80 
 
 85 
 
 
 
 
 
 1 
 
 
 o 
 
 
 
 
 
 
 
 
 3 
 
 
 -1 
 
 85 
 
 90 
 
 
 
 1 
 
 1 
 
 5 
 
 
 
 5 
 
 4 
 
 1 
 
 
 
 
 
 
 23 
 
 
 
 
 90 
 
 95 
 
 
 
 1 
 
 
 
 
 G 
 
 19 
 
 U 
 
 o 
 
 '> 
 
 
 
 
 
 42 
 
 
 1 
 
 95 
 
 100 
 
 
 
 1 
 
 
 o 
 
 y 
 
 8 
 
 11 
 
 5 
 
 
 o 
 
 
 
 
 38 
 
 
 2 
 
 100 
 
 105 
 
 
 
 
 1 
 
 1 
 
 1 
 
 2 
 
 7 
 
 1 
 
 4 
 
 
 2 
 
 
 
 19 
 
 
 3 
 
 105 
 
 110 
 
 
 
 
 
 
 1 
 
 2 
 
 5 
 
 3 
 
 3 
 
 2 
 
 
 1 
 
 
 17 
 
 
 -1 
 
 110 
 
 115 
 
 
 
 
 
 
 
 1 
 
 2 
 
 
 1 
 
 
 2 
 
 1 
 
 
 1 
 
 
 5 
 
 iis'iau 
 
 
 
 
 
 1 
 
 
 
 o 
 
 
 
 
 
 
 
 3 
 
 
 6 
 
 liiO 
 
 125 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 4 
 
 2 
 
 17 
 
 23 
 
 41 
 
 37 
 
 12 
 
 10 
 
 5 
 
 4 
 
 2 
 
 
 158 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 x„ 
 
 -3 
 
 _o 
 
 -1 
 
 
 
 1 
 
 2 
 
 3 
 
 4 
 
 5 
 
 6 
 
 
 
 
 
 
 
 
 28 
 
 29 
 
 30 
 
 31 
 
 32 
 
 33 
 
 34 
 
 35 
 
 36 
 
 37 
 
 
 
 
 X, 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -4 
 
 70 
 
 75 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 -3 
 
 75 
 
 80 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 _o 
 
 80 
 
 85 
 
 
 
 1 
 
 3 
 
 5 
 
 2 
 
 3 
 
 1 
 
 
 
 
 
 15 
 
 
 -1 
 
 85 
 
 90 
 
 
 4 
 
 5 
 
 15 
 
 9 
 
 9 
 
 4 
 
 1 
 
 
 
 
 
 47 
 
 
 
 
 90 
 
 95 
 
 
 1 
 
 7 
 
 S 
 
 8 
 
 8 
 
 3 
 
 O 
 
 2 
 
 
 1 
 
 
 40 
 
 
 1 
 
 95 
 
 100 
 
 
 1 
 
 2 
 
 8 
 
 9 
 
 6 
 
 6 
 
 4 
 
 1 
 
 
 
 
 37 
 
 
 o 
 
 100 
 
 105 
 
 
 
 
 1 
 
 2 
 
 3 
 
 3 
 
 1 
 
 1 
 
 1 
 
 
 
 12 
 
 
 3 
 
 105 
 
 110 
 
 
 
 1 
 
 1 
 
 1 
 
 o 
 
 3 
 
 1 
 
 
 
 
 
 9 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 7 
 
 IG 
 
 36 
 
 34 
 
 30 
 
 22 
 
 10 
 
 4 
 
 1 
 
 1 
 
 
 IGl 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Fig. 17. — Correlation in the buds oV Oeno- 
 thera lamarckiana. No. 04113. Length 
 of ovary subject, thickness of ovary rel- 
 ative. Coefficient of correlation, 0.2724 
 ±0.0492. 
 
 Fig. 16. — Correlation in the buds of Oenothera la- 
 marckiana, 'i^o. 04129. Length of ovary subject, 
 thickness of ovary relative. CoeiBcient of cor- 
 relation, 0.4766 ±0.0522. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 x. 
 
 -11 
 
 -10 
 
 -9 
 
 -8 
 
 —7 
 
 -6 
 
 —5 
 
 -4 
 
 -3 
 
 _o 
 
 -1 
 
 
 
 1 
 
 o 
 
 3 
 
 4 
 
 5 
 
 G 
 
 7 
 
 8 
 
 9 
 
 10 
 
 
 
 
 
 
 
 
 28 
 
 29 
 
 30 
 
 31 
 
 32 
 
 33 
 
 34 
 
 35 
 
 3G 
 
 37 
 
 38 
 
 39 
 
 40 
 
 41 
 
 42 
 
 43 
 
 44 
 
 45 
 
 46 
 
 47 
 
 48 
 
 49 
 
 
 
 
 
 X, 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ■ 
 
 -11 
 
 22 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 2 
 
 
 ■ 
 
 -10 
 
 23 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 -9 
 
 24 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 -8 
 
 25 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 2 
 
 
 
 -7 
 
 2G 
 
 
 
 
 
 
 
 
 
 
 1 
 
 1 
 
 1 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 4 
 
 
 
 -C 
 
 27 
 
 
 
 1 
 
 
 
 
 
 
 
 3 
 
 
 •I 
 
 1 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 8 
 
 
 
 -5 
 
 28 
 
 
 
 
 
 
 
 
 o 
 
 
 2 
 
 
 2 
 
 
 1 
 
 
 1 
 
 1 
 
 
 
 
 
 
 
 
 9 
 
 
 
 -4 
 
 29 
 
 
 
 
 
 
 
 
 2 
 
 
 1 
 
 1 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 5 
 
 
 
 -3 
 
 30 
 
 
 
 
 
 
 1 
 
 
 
 2 
 
 1 
 
 
 1 
 
 o 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 8 
 
 
 
 _o 
 
 31 
 
 
 
 
 
 
 
 
 1 
 
 
 4 
 
 2 
 
 3 
 
 1 
 
 3 
 
 
 
 
 
 
 
 
 
 
 
 14 
 
 
 
 -1 
 
 32 
 
 
 
 
 
 
 
 
 
 
 o 
 
 1 
 
 3 
 
 O 
 
 
 2 
 
 
 1 
 
 
 
 
 
 
 1 
 
 
 12 
 
 
 
 
 
 33 
 
 
 
 
 
 
 
 
 2 
 
 1 
 
 
 1 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 5 
 
 
 
 1 
 
 34 
 
 
 
 
 
 
 
 
 
 3 
 
 
 1 
 
 2 
 
 1 
 
 O 
 
 4 
 
 o 
 
 O 
 
 
 
 
 
 1 
 
 
 
 18 
 
 
 
 o 
 
 35 
 
 
 
 
 
 
 
 
 
 1 
 
 1 
 
 2 
 
 2 
 
 
 4 
 
 3 
 
 o 
 
 
 1 
 
 4 
 
 
 
 
 
 
 20 
 
 
 
 3 
 
 30 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 1 
 
 
 3 
 
 1 
 
 o 
 
 1 
 
 
 
 
 
 
 1 
 
 
 10 
 
 
 
 4 
 
 37 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 O 
 
 o 
 
 
 4 
 
 O 
 
 
 1 
 
 1 
 
 
 
 
 13 
 
 
 
 5 
 
 38 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 o 
 
 1 
 
 2 
 
 2 
 
 1 
 
 3 
 
 3 
 
 1 
 
 1 
 
 
 
 
 17 
 
 
 
 6 
 
 39 
 
 
 
 
 
 
 
 
 
 
 
 
 2 
 
 
 
 
 3 
 
 
 3 
 
 
 
 
 
 
 
 8 
 
 
 
 7 
 
 40 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 o 
 
 1 
 
 
 
 
 
 
 
 3 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 1 
 
 
 
 
 
 1 
 
 1 
 
 I 
 
 7 
 
 16 
 
 11 
 
 19 
 
 10 
 
 20 
 
 IG 
 
 13 
 
 12 
 
 11 
 
 7 
 
 2 
 
 2 
 
 1 
 
 2 
 
 
 IGO 
 
 _ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Fig. 18. — Correlation in buds of Oenothera lamarckiana from Hilversutn. 
 Length of hypanthium subject, length of bud-cone relative. Coefficient of 
 correlation, 0.5797 ± 0.0354. 
 
MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 39 
 
 31 
 n 
 
 8 ^ - 
 
 3 O o 
 
 t n 
 hj 2 
 
 • Q 
 
 
 O K 
 
 3 - 5' 
 2. o 
 
 £. « - 
 ^ ^ c 
 
 5. o- 
 q ?"■ 
 
 £. 5' o 
 =; 3 ° 
 
 1 c ? 
 
 ^ r- ^ 
 ?> — o 
 
 cj re 
 
 ^f^ I 
 
 o -^ § 
 
 o — . ??• 
 
 C*J __. »*. 
 
 ON 2^ O 
 
 OS S a 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 CO 
 
 -J 
 
 
 
 CT 
 
 .£- 
 
 CO 
 
 ro 
 
 - 
 
 c 
 
 _l_ 
 
 1 
 
 10 
 
 1 
 
 CO 
 
 1 
 
 1 
 
 i 
 
 1^ 
 
 1 
 
 1^ 
 
 
 
 
 to 
 
 CO 
 
 X 
 
 
 
 
 
 
 
 j^ 1 j^ 
 
 >^ 
 
 
 
 
 CO 
 
 CO 
 
 SI 
 
 ^ 
 
 CO 
 
 " 
 
 ^ 
 
 — 
 
 CO 
 
 
 
 to 
 
 g 
 
 to 
 
 to 
 
 K 
 
 to 
 
 — 
 
 t-5 
 
 10 
 
 t-3 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 X 
 
 
 
 ^ 
 
 
 
 
 
 
 
 
 
 
 
 
 -- 
 
 
 - 
 
 
 
 
 
 
 
 
 
 t— 
 
 
 ^ 
 
 
 CO 
 tv 
 
 1 
 
 
 
 CO 
 
 
 
 
 
 
 
 
 
 
 
 
 - 
 
 - 
 
 1— 
 
 
 
 
 
 
 
 
 
 
 
 
 
 gl 
 
 1 
 
 a 
 
 
 
 tCi 
 
 
 
 
 
 
 
 
 
 
 - 
 
 
 
 
 
 
 
 
 
 
 >- 
 
 
 
 
 
 
 
 CO 
 
 1 
 
 
 
 CKj 
 
 
 
 
 
 
 
 
 - 
 
 
 
 
 
 
 
 
 - 
 
 to 
 
 to 
 
 
 to 
 
 
 
 
 
 
 
 ^. 
 
 1 
 
 0- 
 
 
 
 tr 
 
 
 
 
 
 
 
 
 
 
 
 
 ►- 
 
 
 - 
 
 - 
 
 CO 
 
 
 - 
 
 
 - 
 
 - 
 
 
 
 
 
 
 ^ 
 
 1 
 
 CO 
 
 
 
 to 
 
 
 
 
 
 ►- 
 
 
 
 - 
 
 CO 
 
 
 CO 
 
 ^ 
 
 l^D 
 
 - 
 
 CO 
 
 CO 
 
 10 
 
 
 to 
 
 
 
 
 
 
 
 
 ^ 
 
 1 
 
 to 
 
 
 
 
 
 
 
 
 
 W» 
 
 1—1 
 
 - 
 
 iO 
 
 CO 
 
 
 
 ii 
 
 - 
 
 ^ 
 
 to 
 
 
 
 - 
 
 
 - 
 
 
 
 
 
 
 CO 
 
 1 
 
 
 
 g 
 
 
 
 
 
 
 
 
 ^ 
 
 
 10 
 
 10 
 
 .;- 
 
 to 
 
 to 
 
 ^ 
 
 - 
 
 
 - 
 
 " 
 
 
 
 
 
 
 
 
 CO 
 
 ^ 
 
 
 
 C-- 
 
 
 
 
 
 
 -- 
 
 H- 
 
 •£- 
 
 IS 
 
 -- 
 
 
 oc 
 
 - 
 
 to 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -- 
 
 
 \Z 
 
 
 
 
 
 
 
 
 CO 
 
 
 CO 
 
 
 to 
 
 - 
 
 
 
 CO 
 
 
 
 
 
 
 
 
 
 
 
 
 LO 
 
 
 
 rt 
 
 
 
 
 - 
 
 
 to 
 
 - 
 
 ^ 
 
 
 
 CO 
 
 J- 
 
 CO 
 
 - 
 
 
 *-' 
 
 
 
 
 
 
 
 
 
 
 
 to 
 
 CO 
 
 
 
 
 
 
 
 
 
 
 
 OS 
 
 
 ts 
 
 
 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 CO 
 
 4. 
 
 
 
 en 
 
 
 
 
 - 
 
 - 
 
 1-3 
 
 
 
 
 
 
 " 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 t^ 
 
 cn 
 
 
 
 c; 
 
 
 ^ 
 
 
 
 
 
 - 
 
 
 " 
 
 - 
 
 
 - 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 en 
 
 cs 
 
 
 
 ^ 
 
 
 
 
 
 - 
 
 
 
 
 - 
 
 - 
 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 (?> 
 
 --I 
 
 
 
 OJ 
 
 
 
 
 
 - 
 
 
 
 
 - 
 
 
 
 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -^ 
 
 cc 
 
 
 
 CO 
 
 
 
 
 - 
 
 
 - 
 
 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 !S 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 i 
 
 
 H- 
 
 
 
 &3 
 
 ■^ 
 
 -^ 
 
 •^ 
 
 &I t>^ L-w 
 
 cc g 
 
 ^ »— 
 
 -'{■-.i-^ ■^ 
 
 ^ 
 
 *- 
 
 to 
 
 
 
 ^ 
 
 
 
 H- 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 1 
 
 
 
 
 
 
 
 
 
 
 
 
 o c' s- 
 
 p 
 
 O 
 
 H-S. 
 
 p r* 
 
 c^ ^ a 
 <^ C- 
 
 2; 
 ' o 
 
 o V ■ 
 
 • e 
 
 O D. i ^ 
 
 n 
 
 "^ 3' 
 
 ■r f 
 
 ^ CD *^ 
 
 -•3 2* 
 < jt) p 
 
 P Jo- 
 
 O O I 
 o "• 
 
 2.§ o 
 
 2 » <^ 
 
 3 
 
 - c 5 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 t— 
 
 
 
 
 
 OD 
 
 ^1 
 
 
 
 01 
 
 •c>. 
 
 CO 
 
 10 
 
 - 
 
 c 
 
 \^ 
 
 .!, 
 
 1 
 
 CO 
 
 1 
 
 lb 
 
 1 
 
 en 
 
 1 
 
 C5 
 
 I 
 
 1 
 
 00 
 
 ; 
 
 X 
 
 
 
 
 
 
 
 
 
 CO 
 
 D3 
 
 CO 
 
 CO 
 •'I 
 
 s? 
 
 ^ 
 
 CO 
 
 lb 
 
 CO 
 
 CO 
 
 CO 
 
 to 
 
 CO 
 
 CO 
 
 
 
 to 
 
 to 
 
 00 
 
 I^ 
 
 to 
 
 CI 
 
 to 1 to 
 
 V* 1 lb 
 
 to 
 
 CO 
 
 n 
 
 to 
 
 '£ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 X 
 
 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -- 
 
 
 
 
 
 
 g 
 
 1 
 
 ^4 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 &9 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 CO 
 
 to 
 
 1 
 01 
 
 
 
 cc 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 h- 
 
 to 
 
 
 
 
 
 
 
 
 CO 
 
 CO 
 
 1 
 
 lb 
 
 
 
 c 
 
 
 
 
 
 
 
 
 !-• 
 
 
 
 
 
 
 
 CO 
 
 to 
 
 
 to 
 
 - 
 
 - 
 
 
 
 
 CO 
 
 1 
 
 CO 
 
 
 
 CO 
 
 
 
 
 
 
 
 
 to 
 
 
 
 '^ 
 
 - 
 
 to 
 
 to 
 
 CT 
 
 
 to 
 
 C.1 
 
 " 
 
 
 to 
 
 
 
 CO 
 
 1 
 
 to 
 
 
 
 -I 
 
 
 
 
 
 
 l-» 
 
 - 
 
 ►- 
 
 
 to 
 
 - 
 
 - 
 
 - 
 
 - 
 
 »-• 
 
 1- 
 
 t3 
 
 t-l 
 
 
 - 
 
 
 to 
 
 
 CO 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 - 
 
 
 .£. 
 
 to 
 
 to 
 
 
 - 
 
 t: 
 
 -J 
 
 ►- 
 
 »-» 
 
 " 
 
 
 
 
 
 CO 
 
 c 
 
 
 
 to 
 
 
 
 
 
 
 - 
 
 
 to 
 
 to 
 
 to 
 
 l-» 
 
 CO 
 
 to 
 
 ^"^ 
 
 to 
 
 
 
 ts 
 
 - 
 
 
 ►- 
 
 
 
 CO 
 
 OD 
 
 l-l 
 
 
 
 Li 
 
 
 
 
 
 - 
 
 
 CO 
 
 - 
 
 
 10 
 
 00 
 
 to 
 
 t9 
 
 CO 
 
 
 ►- 
 
 CO 
 
 
 
 
 
 
 
 s 
 
 to 
 
 
 
 -1 
 
 
 
 
 
 
 
 - 
 
 to 
 
 CO 
 
 ►- 
 
 - 
 
 H. 
 
 ►- 
 
 CO 
 
 « 
 
 to 
 
 
 
 h- 
 
 
 
 
 
 c 
 
 CO 
 
 
 
 c- 
 
 
 
 
 ^ 
 
 
 
 to 
 
 
 
 
 - 
 
 - 
 
 
 
 H- 
 
 
 
 
 
 
 
 
 
 •u 
 
 lb 
 
 
 
 Ci 
 
 
 
 
 
 
 
 to 
 
 
 - 
 
 
 
 to 
 
 
 
 
 
 
 
 
 
 
 
 
 to 
 
 en 
 
 
 
 cc 
 
 
 
 - 
 
 
 - 
 
 
 
 
 to 
 
 CO 
 
 
 to 
 
 
 
 
 
 
 
 
 
 
 
 
 CO 
 
 c; 
 
 
 
 CO 
 
 
 - 
 
 
 
 - 
 
 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 .u 
 
 ^t 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 C-. 
 
 
 - 
 
 ^ 
 
 
 CO 
 
 ti 
 
 = 
 
 ^ 
 
 ex 
 
 I21 
 
 
 
 OT 
 
 CC 
 
 ^T 
 
 05 
 
 s> 
 
 ex 
 
 I-* 
 
 IC 
 
 CT 
 
 to 
 
 CO 
 
 to 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
40 MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE) OENOTHERAS. 
 
 
 
 
 
 
 
 
 
 
 
 1 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 •M 
 
 eo 
 
 C) 
 
 ira 
 
 
 CO 
 
 o 
 
 o i o 
 
 CO 
 
 
 o 
 
 ■S" 
 
 CO 
 
 n 
 
 :rs 
 
 M 
 
 CO 
 
 - 
 
 - 
 
 -^ 
 
 
 o 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 uTt 
 
 o 
 
 
 
 
 
 
 
 
 
 
 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 -- 
 
 
 
 •V 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 - 
 
 
 
 
 
 
 
 
 
 
 -■ 
 
 
 
 CO 
 
 ?? 
 
 
 
 
 
 ^ 
 
 ^^ 
 
 
 - 
 
 ^^ 
 
 - 
 
 
 c^ 
 
 - 
 
 
 (N 
 
 
 1-* 
 
 
 
 
 
 
 
 ::; 
 
 
 
 C-1 
 
 
 
 
 - 
 
 '"' 
 
 
 --; 
 
 
 - 
 
 -r ' ?l 
 
 n 
 
 - 
 
 
 - 
 
 CO 
 
 -- 
 
 
 eo 
 
 
 
 
 
 
 
 
 1-1 
 
 
 
 
 
 
 -H 
 
 ci ! CO 
 
 -r -^ ci — ' 
 
 -T 
 
 CO 
 
 -r ! CO 
 
 CI ' -T 
 
 
 
 
 
 ^^ 
 
 
 ? 
 
 
 
 o 
 
 -^ 
 
 
 C-1 
 
 -^ 
 
 
 ^ 
 
 •^ 
 
 CO 
 
 •^ 
 
 — CO 
 
 CO 
 
 *r 
 
 lo 
 
 o 
 
 CI 
 
 OJ 
 
 CO 
 
 
 
 »H 
 
 -H 
 
 
 
 
 
 
 -H 
 1 
 
 
 
 
 f-( 
 
 - 
 
 -^ 
 
 M 
 
 -^ 
 
 o 
 
 CI rr 
 
 " 
 
 - 
 
 
 
 
 iH 
 
 
 t-t 
 
 
 
 
 
 
 CO 
 
 
 
 ! 
 
 ?! 
 
 
 
 
 
 ^ 
 
 - 
 
 .-1 
 
 
 N 
 
 - 
 
 - 
 
 - 
 
 
 ?i 
 
 
 i-t 
 
 
 - 
 
 
 
 
 
 
 r^j 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 (M 
 
 
 
 
 
 
 
 
 c^i 
 
 
 
 X 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 8 
 
 c-1 
 
 f} 
 
 
 ^ 
 
 
 ?i 
 
 i-1 
 
 r? 1 5i 
 
 o 
 
 CO 
 
 CO 
 
 CO 
 
 CO 
 
 s 
 
 S 
 
 55 
 
 S 
 
 S 
 
 ? 
 
 
 
 
 
 
 
 X 
 
 ^ 
 1 
 
 00 
 
 1 
 
 1 
 
 to 
 
 1 
 
 1/5 
 
 1 
 
 I 
 
 CO 
 
 I 
 
 CJ 
 
 T ° 
 
 — ca 
 
 CO 
 
 T 
 
 *o 
 
 CD 
 
 r* 
 
 CO 
 
 cri 
 
 o 
 
 ^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ;^ 
 
 >. 
 
 g 
 
 "^j 
 
 /"^ 
 
 rt 
 
 ■::^ 
 
 <M 
 
 C 
 
 ;2 
 
 
 
 r^ 
 
 
 
 
 n 
 
 •jf 
 
 t*- 
 
 
 M 
 
 o 
 
 
 
 
 
 
 o 
 
 tfl 
 
 
 hJ 
 
 o 
 
 
 
 
 3 
 
 , 
 
 ^ 
 
 o 
 
 o 
 
 :2: .a- 
 
 ■B^-a ^^ 
 
 ^ ^ Vi 
 
 C 
 
 o 
 
 = ■5 2 
 
 5 '^ s „- 
 
 
 
 n 
 
 
 n 
 
 
 ~ 
 
 
 
 
 
 
 1 — 
 
 
 
 
 
 
 
 
 
 
 
 
 
 I 1 
 
 
 
 
 
 - 
 
 O I -1 
 
 o 
 
 n 
 
 -r 
 
 -r 
 
 -r 
 
 -1* 
 
 M 
 
 t^ : — 
 
 h- 
 
 ?? 
 
 <» 2 
 
 CO 
 
 IC 
 
 t^ 
 
 r^ 
 
 -r 
 
 eo 
 
 o 
 
 -• 
 
 CO 
 
 1 i2 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 i 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ^ 
 
 CI 
 
 
 
 
 
 
 
 
 
 
 
 
 - 
 
 1-* 
 
 -• 
 
 
 
 
 
 
 
 
 
 
 
 
 
 CO 
 
 
 
 CO 
 
 CO 
 CI 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1-i 
 
 
 •-« 
 
 
 
 
 
 
 .-1 
 
 
 
 
 
 
 CO 
 
 
 
 ^1 
 
 s 
 
 
 
 
 
 
 
 
 -. 
 
 
 
 
 
 -H 
 
 eo 
 
 - 
 
 
 iH 
 
 ■o 
 
 t-t 
 
 -- 
 
 C5 
 
 -< 
 
 
 
 
 I^ 
 
 
 
 -H 
 
 s 
 
 
 
 
 -H 
 
 
 - 
 
 - 
 
 
 -H 
 
 _ 
 
 CO 
 
 CI 
 
 CI ■ CO !ir 
 
 C, Ti^Uo 
 
 eo 
 
 CO 
 
 - 
 
 
 
 -• 
 
 
 
 
 
 o 
 
 s 
 
 
 
 
 
 
 - 
 
 •-4 
 
 CO 
 
 CI 
 
 
 CI 
 
 CO 
 
 ci — a: 
 
 — 1 -r i CO 
 
 ^ 
 
 Cl 
 
 CI 
 
 
 - 
 
 
 
 
 CO 
 
 
 
 1 
 
 c5 
 
 
 
 
 
 
 
 -H 
 
 
 
 CJ 
 
 - 
 
 
 CI 
 
 eo 1 — 
 
 C^l 
 
 CO 
 
 CO 
 
 -r 
 
 - 
 
 
 
 -H 
 
 
 
 
 CI 
 
 
 
 1 
 
 ?, 
 
 
 ^H 
 
 
 
 
 
 - 
 
 
 -H 
 
 - 
 
 CI 
 
 - 
 
 CI 
 
 CO 
 
 CO 
 
 
 
 
 
 
 
 - 
 
 
 
 
 
 cs 
 
 
 
 X 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 s 
 
 s 
 
 S3 
 
 ?3 
 
 S 
 
 
 8 
 
 S 
 
 s 
 
 C5 
 CJ 
 
 § 
 
 eo 
 
 CI 
 
 CO 
 
 SI 
 
 eo 
 
 eo 
 
 s 
 
 ^ 
 
 00 
 
 CO 
 
 ? 
 
 
 C-1 
 T 
 
 eo 
 
 •r 
 
 
 
 
 
 
 
 X 
 
 2 
 
 fj 
 
 ^ 
 
 2 
 
 C5 
 1 
 
 00 
 
 1 
 
 1 
 
 CO 
 
 1 
 
 1 
 
 -1« 
 1 
 
 eo 
 
 1 
 
 c-i 
 
 1 
 
 T 
 
 o 
 
 - 
 
 M 
 
 CO 
 
 -r 
 
 LO 
 
 o 
 
 t- 
 
 00 
 
 cs 
 
 o 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 L 
 
 
 
 
 
 
 
 
 J 
 
 J 
 
 
 
 
 
 
 
 
 
 ^ 2 = 
 
 ta ^ o 
 
 ?■ J2 '5 '^" 
 
 1-^ .ii 'r- >o 
 
 
 c\ _ 
 
 — o 
 
 •S s 1 1 s 
 
 ^ O « rJ — 
 
 .£ ^. 3 -5 O 
 
 ti c s = 
 
 •3 I := 5 § 
 
 ■Sil^ 
 
 gc, 
 
 tN 
 
 >'8 
 
 O c c o 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 CI 
 
 - 
 
 .-1 
 
 CI 
 
 *T 
 
 CO 
 
 C3 
 
 lO 
 
 00 
 
 •H 
 
 
 i-O 
 
 1-t 
 
 s 
 
 o , eo 
 
 r* 
 
 00 
 
 eo 
 
 
 s 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ■^ 
 
 S 
 
 
 
 
 
 
 
 
 
 
 
 
 1-1 
 
 
 
 
 
 
 
 
 
 
 - 
 
 
 
 CO 
 
 CI 
 
 
 
 
 
 
 
 
 
 
 
 iH 
 
 
 
 
 
 
 
 
 
 
 
 1-r 
 
 
 
 c^ 
 
 CO 
 CI 
 
 
 
 
 
 
 
 .H 
 
 ^ 
 
 ^ 
 
 - 
 
 
 c-l 
 
 
 CI 
 
 CO 
 
 --* 
 
 Cl 
 
 1-H 
 
 i-H 
 
 
 
 o 
 
 
 
 ^ 
 
 r-- 
 
 
 
 
 i-H 
 
 *-l 
 
 c-l 
 
 eo 
 
 c-l 
 
 eo 
 
 lO 
 
 CO 
 
 eo 
 
 CJ 
 
 M 
 
 ■r ' LO 
 
 Cl 
 
 o 
 
 Cl 
 
 C-l 
 
 
 !^! 
 
 
 o 
 
 CI 
 
 
 ^ 
 
 
 
 ^H 
 
 ^^ 
 
 M 
 
 •r 
 
 1-t 
 
 
 T 
 
 CO 
 
 ^H 
 
 o 
 
 o 
 
 T 
 
 LO 
 
 3i 
 
 r-t 
 
 .-« 
 
 
 -r 
 
 
 
 t-H 
 
 ! 
 
 lO 
 C) 
 
 
 
 
 
 
 
 T-t 
 
 -H 
 
 
 .-t 
 
 i-O 
 
 C^l 
 
 
 t-O 
 
 LO 
 
 
 eo 
 
 
 ■V 
 
 
 
 c« 
 
 
 
 c^ 
 
 1 
 
 -r 
 CI 
 
 
 
 
 
 
 
 1-1 
 
 ^ 
 
 
 l-< 
 
 1-« 
 
 •-4 
 
 - 
 
 c-l 
 
 CI 
 
 
 
 (-1 
 
 
 
 
 ^H 
 
 
 
 eo 
 1 
 
 CO 
 CI 
 
 
 i-H 
 
 
 
 
 rH 
 
 
 
 
 
 
 
 1-1 
 
 r-l 
 
 -. 
 
 
 
 ^H 
 
 
 
 
 o 
 
 
 
 1 
 
 CI 
 
 
 
 -^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 f-H 
 
 
 
 
 
 cl 
 
 
 
 b<' 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ?! 
 
 ?5 
 
 c> 
 
 to 
 c^ 
 
 s 
 
 c^ 
 
 CO 
 CI 
 
 ?s 
 
 s 
 
 eo 
 
 5? 
 
 eo 
 eo 
 
 CO 
 
 CO 
 
 CD 
 CO 
 
 co 
 
 00 
 CO 
 
 S 
 
 ? 
 
 
 
 
 
 
 
 X 
 
 z* 
 
 o 
 
 Ci 
 
 1 
 
 00 
 
 1 
 
 r 
 
 1 
 
 lO 
 
 1 
 
 CO 
 
 c^ 
 
 1 
 
 T 
 
 o 
 
 -- 
 
 CI 
 
 CO 
 
 -r 
 
 o 
 
 cs 
 
 I-- 
 
 
 
 
 
 
 
 
 
 
 u 
 
 _ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ^ c ^ o 
 
 a V o u 
 
 I ^"^ 
 
 O a; ^ 
 
 "s s ll 
 
 
 -1 
 
 .X. 
 
 '■J 
 
 
 T/1 
 
 t;^ 
 
 " 
 
 
 
 TJ 
 
 u 
 
 -*-» 
 
 
 -o 
 
 > 
 
 ■4-» 
 
 
 
 CO 
 
 
 (— t 
 
 o 
 
 
 lO 
 
 c 
 
 HI 
 
 Si 
 
 
 o 
 
 c 
 
 
 
 c 
 
 3 
 
 > 
 
 
 -*-• 
 
 ^-1 
 
 c 
 
 
 •* 
 
 
 
 y 
 
 
 ^ 
 
 o 
 
 
 
 
 )-< 
 
 •* 
 o 
 
 u 
 
 " 
 
 
 71 
 
 
 
 
 
 G 
 
 3 
 
 o 
 
 1 
 
 »>. 
 
 >, 
 
 _^ 
 
 j3 
 
 
 Cl 
 
 
 •«-> 
 
 
 (N 
 
 5 
 
 "o 
 
 5 
 
 
 
 
 
 
 C 
 
 
 
 
 
 
 
 
 
 
 Ul 
 
 
 
 
 
MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 4 1 
 
 
 
 
 X.}-2 
 
 -1 
 
 
 
 1 
 
 
 
 •J 
 
 4 
 
 5 
 
 
 
 
 
 
 
 
 60 
 
 70 
 
 75 
 
 80 
 
 85 
 
 90 
 
 95 100i 
 
 
 
 
 
 
 
 70 
 
 75 
 
 SO 
 
 85 
 
 90 
 
 65 
 
 100,105 
 
 
 
 
 
 X, 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -11 
 
 28 
 
 
 1 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 -10 
 
 29 
 
 
 1 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 -9 
 
 30 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -s 
 
 31 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ~ t 
 
 32 
 
 
 
 1 
 
 
 
 
 
 
 
 
 1 
 
 
 
 -G 
 
 33 
 
 
 1 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 -5 
 
 34 
 
 
 
 2 
 
 3 
 
 1 
 
 1 
 
 
 
 
 
 7 
 
 
 
 -4 
 
 35 
 
 
 
 1 
 
 4 
 
 
 
 
 
 
 
 
 7 
 
 
 
 _o 
 
 3G 
 
 
 1 
 
 2 
 
 11 
 
 2 
 
 
 
 
 
 
 IG 
 
 
 
 -0 
 
 37 
 
 
 1 
 
 
 
 
 
 7 
 
 
 
 
 
 
 11 
 
 
 
 -1 
 
 3sf 
 
 
 
 2 
 
 5 
 
 10 
 
 1 
 
 1 
 
 
 
 
 li) 
 
 
 
 
 
 39 
 
 
 
 
 2 
 
 5 
 
 3 
 
 
 
 
 
 10 
 
 
 
 1 
 
 40 
 
 
 
 2 
 
 i 
 
 G 
 
 5 
 
 
 
 
 
 20 
 
 
 
 2 
 
 41 
 
 
 
 
 4 
 
 5 
 
 4 
 
 3 
 
 
 
 
 u; 
 
 
 
 3 
 
 42 
 
 
 
 
 
 G 
 
 C 
 
 
 1 
 
 
 
 13 
 
 
 
 4 
 
 13 
 
 
 
 
 
 4 
 
 i 
 
 1 
 
 
 
 
 12 
 
 
 
 5 
 
 44 
 
 
 
 
 
 4 
 
 4 
 
 2 
 
 1 
 
 
 
 11 
 
 
 
 G 
 
 45 
 
 
 
 
 
 3 
 
 
 
 2 
 
 
 
 
 i 
 
 
 
 7 
 
 •3 
 
 
 
 
 
 
 *> 
 
 
 
 
 
 2 
 
 
 
 8 
 
 47 
 
 
 
 
 
 
 1 
 
 
 
 1 
 
 
 2 
 
 
 
 9 
 
 ii 
 
 
 
 
 
 
 1 
 
 
 
 
 
 1 
 
 
 
 10 
 
 4b 
 
 
 
 
 
 
 1 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 5 
 
 10 
 
 39 
 
 55! 33 
 
 10 
 
 
 
 1 
 
 iGO, 
 
 
 
 —^ ]-' 
 
 
 i ' 
 
 
 
 
 
 
 
 
 Xo 
 
 -4 
 
 ~3 
 
 _o 
 
 -I 
 
 
 
 1 
 
 2 
 
 3 
 
 4 
 
 5 
 
 6 
 
 
 *"" 
 
 
 
 
 
 
 CO 
 
 65 
 
 70 
 
 75 
 
 80 
 
 85 
 
 90 
 
 95 100 
 
 105 
 
 UO 
 
 
 
 
 
 
 
 
 05 
 
 70 
 
 75 
 
 80 
 
 85 
 
 90 
 
 95 
 
 lOO'lCS 
 
 110!ll5 
 
 
 
 
 Xi 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ~ / 
 
 32 
 
 
 1 
 
 
 2 
 
 1 
 
 
 
 
 
 
 
 
 
 4 
 
 
 
 -6 
 
 33 
 
 
 
 
 
 2 
 
 1 
 
 
 
 
 
 
 
 
 3 
 
 
 
 ■~5 
 
 31 
 
 
 
 
 
 2 
 
 
 
 
 
 
 
 
 
 2 
 
 
 ■ 
 
 -4 
 
 35 
 
 
 
 
 2 
 
 4 
 
 2 
 
 
 
 
 
 
 
 
 8 
 
 
 
 -3 
 
 3C 
 
 
 
 
 1 
 
 G 
 
 2 
 
 
 
 
 
 
 
 
 9 
 
 
 
 _2 
 
 o7 
 
 
 
 
 3 
 
 10 
 
 G 
 
 3 
 
 
 
 
 
 
 
 »>■> 
 
 
 
 -1 
 
 38 
 
 
 
 
 
 8 
 
 G 
 
 4 
 
 1 
 
 
 
 
 
 
 19 
 
 
 
 
 39 
 
 
 
 
 
 3 
 
 '.> 
 
 7 
 
 1 
 
 
 
 
 
 
 20 
 
 
 
 1 
 
 40 
 
 
 
 
 
 
 6 
 
 8 
 
 2 
 
 
 
 
 
 
 16 
 
 
 
 
 
 41 
 
 
 
 
 
 
 3 
 
 5 
 
 3 
 
 
 
 
 
 
 11 
 
 
 
 _3j 
 
 42 
 
 
 
 
 
 
 1 
 
 G 
 
 9 
 
 1 
 
 
 
 
 
 17 
 
 
 
 4 
 
 43 
 
 
 
 
 
 
 
 1 
 
 5 
 
 
 
 
 
 
 G 
 
 
 
 5 
 
 44 
 
 
 
 
 
 
 
 
 4 
 
 1 
 
 
 
 
 
 5 
 
 
 
 C 
 
 45 
 
 
 
 
 
 
 
 
 3 
 
 3 
 
 
 
 
 
 G 
 
 
 
 7 
 
 46 
 
 
 
 
 
 
 
 
 2 
 
 2 
 
 
 
 
 
 4 
 
 
 
 8 
 
 47 
 
 
 
 
 
 
 
 
 
 1 
 
 2 
 
 
 
 
 3 
 
 
 
 9 
 
 48 
 
 
 
 
 
 
 
 
 
 1 
 
 1 
 
 
 1 
 
 
 3 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 8 
 
 36 
 
 36 
 
 34 
 
 30 
 
 9 
 
 3 
 
 
 
 1 
 
 
 158 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Fig. 24. — -Correlation in buds of O. lamarcki- 
 ana from Hilversum. Lengtli of bud- 
 cone subject, thickness of bud-cone 
 relative. Coefficient of correlation, 
 0.7852 ± 0.0204. 
 
 Fig. 25. — Correlation in the buds of Oenothera 
 lamarckiana. No. 04129. Length of bud-cone 
 subject, thickness of bud-cone relative. Coeffi- 
 cient of correlation, 0.8732 ± 0.0127. 
 
 
 
 
 Xo 
 
 -4 
 
 f^ 
 
 -2I-I 
 
 
 
 1 
 
 2 3 
 
 11 
 
 
 
 
 
 
 C5 
 
 70 
 
 75 
 
 SO 
 
 S5 
 
 90 
 
 95 100 
 
 
 
 
 
 
 
 
 70 
 
 75 
 
 80 
 
 85 
 
 90 
 
 95 
 
 '100105 
 
 
 
 
 
 Xi 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ~ t 
 
 30 
 
 
 
 
 
 1 
 
 
 
 
 
 
 1 
 
 
 
 -6 
 
 31 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -5 
 
 32 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -4 
 
 33 
 
 
 1 
 
 1 
 
 
 1 
 
 
 
 
 
 
 3 
 
 
 
 -3 
 
 34 
 
 
 
 1 
 
 4 
 
 5 
 
 
 
 
 
 
 10 
 
 
 
 -2 
 
 35 
 
 
 
 
 6 
 
 12 
 
 4 
 
 1 
 
 
 
 
 23 
 
 
 
 -1 
 
 36 
 
 
 
 
 3 
 
 6 
 
 6 
 
 2 
 
 
 
 
 17 
 
 
 
 
 
 37 
 
 
 
 
 1 
 
 4 
 
 13 
 
 4 
 
 
 
 
 22 
 
 
 
 1 
 
 38 
 
 
 
 
 
 2 
 
 13 
 
 8 
 
 1 
 
 
 
 24 
 
 
 
 2 
 
 39 
 
 
 
 
 
 1 
 
 7 
 
 10 
 
 2 
 
 1 
 
 
 21 
 
 — 
 
 
 3 
 
 40 
 
 
 
 
 
 
 5 
 
 9 
 
 3 
 
 
 
 17 
 
 
 4 
 
 41 
 
 
 
 
 
 
 
 2 
 
 4 
 
 
 
 6 
 
 
 
 5 
 
 42 
 
 
 
 
 
 
 
 
 5 
 
 
 
 5 
 
 
 
 6 
 
 43 
 
 
 
 
 
 
 
 1 
 
 7 
 
 1 
 
 
 9 
 
 
 
 7 
 
 44 
 
 
 
 
 
 
 
 1 
 
 
 2 
 
 
 3 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 2 
 
 14 
 
 32 
 
 48 
 
 38 
 
 22 
 
 4 
 
 
 161 
 
 
 
 
 
 _ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 x. 
 
 -4 
 
 -3 
 
 _2 
 
 -1 
 
 
 
 1 
 
 2 
 
 3 
 
 4 
 
 
 
 
 
 
 
 
 22 
 
 23 
 
 24 
 
 25 
 
 2G 
 
 27 
 
 28 
 
 29 
 
 30 
 
 
 
 
 
 Xi 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 —7 
 
 2G 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 -6 
 
 27 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 1 
 
 
 
 ~5 
 
 2« 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 1 
 
 
 
 -4 
 
 29 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 -3 
 
 30 
 
 
 1 
 
 
 2 
 
 3 
 
 1 
 
 
 
 
 
 
 7 
 
 
 
 —2 
 
 31 
 
 
 
 1 
 
 1 
 
 1 
 
 3 
 
 1 
 
 
 
 
 
 7 
 
 
 
 -1 
 
 32 
 
 
 
 2 
 
 4 
 
 11 
 
 11 
 
 5 
 
 3 
 
 
 
 
 36 
 
 
 
 
 
 33 
 
 
 
 1 
 
 
 3 
 
 12 
 
 G 
 
 2 
 
 
 
 
 24 
 
 
 
 1 
 
 34 
 
 
 
 
 2 
 
 6 
 
 11 
 
 11 
 
 2 
 
 
 
 
 32 
 
 
 
 
 
 35 
 
 
 
 1 
 
 
 1 
 
 7 
 
 13 
 
 
 1 
 
 
 
 23 
 
 
 
 3 
 
 30 
 
 
 
 
 
 1 
 
 2 
 
 8 
 
 3 
 
 
 1 
 
 
 15 
 
 
 
 4 
 
 37 
 
 
 
 
 
 1 
 
 1 
 
 3 
 
 5 
 
 
 
 
 10 
 
 
 
 5 
 
 38 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 1 
 
 
 
 6 
 
 39 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 7 
 
 40 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 6 
 
 11 
 
 27 
 
 49 
 
 47 
 
 16 
 
 1 
 
 1 
 
 
 160 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Fig. 26. — Correlation in the bud of Oeno- 
 thera lamarckiana. No. 04113. Length 
 of bud-cone subject, thickness of bud- 
 cone relative. Coefficient of correla- 
 tion, 0.8289 ±0.0166. 
 
 Fig. 27. — Correlation in the buds of Oenothera 
 lamarckiana from Hilversum. Thickness 
 of ovary subject, thickness of hypanthium . 
 relative. Coefficientof correlation, 0.5727 
 ±0.0358. 
 
42 MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 x. 
 
 — O 
 
 -1 
 
 
 
 1 
 
 2 
 
 3 
 
 4 
 
 
 
 
 
 
 
 
 23 
 
 24 
 
 25 
 
 26 
 
 27 
 
 28 
 
 29 
 
 
 
 
 
 Xi 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -5 
 
 27 
 
 
 1 
 
 
 
 
 
 
 
 
 1 
 
 
 
 -4 
 
 28 
 
 
 1 
 
 3 
 
 
 
 
 
 
 
 4 
 
 
 
 -3 
 
 29 
 
 
 
 1 
 
 
 
 
 
 1 
 
 
 2 
 
 
 
 _2 
 
 30 
 
 
 5 
 
 5 
 
 5 
 
 2 
 
 
 
 
 
 17 
 
 
 
 -1 
 
 31 
 
 
 7 
 
 4 
 
 G 
 
 4 
 
 2 
 
 
 
 
 23 
 
 
 
 
 
 32 
 
 
 2 
 
 G 
 
 22 
 
 8 1 3 
 
 
 
 
 41 
 
 
 
 1 
 
 33 
 
 
 
 O 
 
 10 
 
 21 
 
 4 
 
 
 
 
 37 
 
 
 
 2 
 
 34 
 
 
 
 o 
 
 id 
 
 4 
 
 3 
 
 3 
 
 
 
 
 12 
 
 
 
 3 
 
 35 
 
 
 
 
 
 5 
 
 1 
 
 2 
 
 2 
 
 
 10 
 
 
 
 4 
 
 36 
 
 
 
 1 
 
 
 3 
 
 1 
 
 
 
 
 5 
 
 
 
 5 
 
 37 
 
 
 
 
 1 
 
 1 
 
 2 
 
 
 
 
 4 
 
 
 
 6 
 
 38 
 
 
 
 
 
 
 1 
 
 1 
 
 
 
 2 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 16 
 
 24 
 
 48 
 
 47 
 
 17 
 
 3 
 
 3 
 
 
 158 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Fig. 28. — Correlation in the bitds of O. 
 lamarckiana, 0.4129. Thickness of 
 ovary subject, thickness of hypan- 
 thium relative. Coefficient of cor- 
 relation, 0.5579 ± 0.0370. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 Xo 
 
 -3 
 
 —o 
 
 -1 
 
 
 
 1 
 
 2 
 
 3 
 
 4 
 
 5 
 
 
 
 
 
 
 
 
 22 
 
 23 
 
 24 
 
 25 
 
 26 
 
 27 
 
 28 
 
 29 
 
 30 
 
 
 
 
 
 Xi 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -3,28 
 
 
 
 o 
 
 3 
 
 1 
 
 1 
 
 
 
 
 
 
 7 
 
 
 
 — n 
 
 29 
 
 
 1 
 
 3 
 
 3 
 
 3 
 
 6 
 
 
 
 
 
 
 16 
 
 
 
 -1 
 
 30 
 
 
 1 
 
 4 
 
 9 
 
 13 
 
 6 
 
 2 
 
 1 
 
 
 
 
 36 
 
 
 
 
 
 31 
 
 
 
 2 
 
 G 
 
 16 
 
 5 
 
 5 
 
 
 
 
 
 34 
 
 
 
 1 
 
 32 
 
 
 
 
 o 
 
 5 
 
 12 
 
 10 
 
 1 
 
 
 
 
 30 
 
 
 
 2 
 
 33 
 
 
 
 1 
 
 
 9 
 
 6 
 
 4 
 
 1 
 
 1 
 
 
 
 oo 
 
 
 
 3 
 
 34 
 
 
 
 
 
 2 
 
 1 
 
 1 
 
 6 
 
 
 
 
 10 
 
 
 
 4 
 
 35 
 
 
 
 
 
 
 2 
 
 
 1 
 
 
 1 
 
 
 4 
 
 
 
 5 
 
 36 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 1 
 
 
 
 6 
 
 37 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 2 
 
 12 
 
 23 
 
 49 
 
 40 
 
 22 
 
 11 
 
 1 
 
 1 
 
 
 161 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Fig. 29. — Correlation in the buds of Oeno- 
 thera lamarckiana, Xo. 041 13. Thickness 
 of ovary subject, thickness of hypan- 
 thium relative. Coefficient of correlation, 
 0.5565 + 0.0367. 
 
 
 
 
 Xj-2 
 
 -1 
 
 
 
 1 
 
 2 
 
 3|4 
 
 5 
 
 
 
 
 
 
 
 
 65 
 
 70 
 
 75 
 
 80 
 
 85 
 
 90 
 
 95 
 
 100 
 
 
 
 
 
 
 
 
 70 
 
 75 
 
 80 
 
 85 
 
 90 
 
 95 
 
 100 
 
 105 
 
 
 
 
 
 X, 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -7 
 
 2G 
 
 
 1 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 -6 
 
 27 
 
 
 1 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 ~5 
 
 28 
 
 
 
 1 
 
 
 
 
 
 
 
 
 1 
 
 
 
 -4'2U 
 
 
 1 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 -3 30 
 
 
 1 
 
 2 
 
 3 
 
 1 
 
 
 
 
 
 
 7 
 
 
 
 -2; 31 
 
 
 1 
 
 1 
 
 o 
 
 3 
 
 
 
 
 
 
 7 
 
 
 
 -li32 
 
 
 
 4 
 
 8 
 
 17 
 
 C 
 
 1 
 
 
 
 
 36 
 
 
 
 J33 
 
 
 
 
 9 
 
 7 
 
 6 
 
 o 
 
 
 
 
 24 
 
 
 
 1 |34 
 
 
 
 2 
 
 11 
 
 10 
 
 9 
 
 
 
 
 
 32 
 
 
 
 2 35 
 
 
 
 
 4 
 
 9 
 
 G 
 
 4 
 
 
 
 
 23 
 
 
 
 3 
 
 36 
 
 
 
 
 o 
 
 5 
 
 4 
 
 3 
 
 1 
 
 
 
 15 
 
 
 
 4 
 
 37 
 
 
 
 
 
 3 
 
 5 
 
 
 1 
 
 1 
 
 
 10 
 
 
 
 5 
 
 38 
 
 
 
 
 
 
 1 
 
 
 
 
 
 1 
 
 
 
 C 
 
 39 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 7 
 
 40 
 
 
 
 
 
 
 1 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 5 
 
 10 
 
 39 
 
 55 
 
 38 
 
 10 
 
 2[1 
 
 
 160 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Fig. 30. — Correlation in buds of Oenothera 
 lamarckianairom Hilversum. Thick- 
 ness of ovary subject, thickness of bud- 
 cone relative. Coefficient of correla- 
 tion, 0.5691 ± 0.0360. 
 
 
 
 
 Xo 
 
 -4 
 
 -3 
 
 _o 
 
 -1 
 
 
 
 1 
 
 O 
 
 3 
 
 4 
 
 5 
 
 6 
 
 
 
 
 
 
 
 
 60 
 
 65 
 
 70 
 
 75 
 
 80 
 
 85 
 
 90 
 
 95 
 
 100 
 
 105 
 
 110 
 
 
 
 
 
 
 
 
 65 
 
 70 
 
 75 
 
 80 
 
 85 
 
 90 
 
 95 
 
 lOOJlOS 
 
 110115 
 
 
 
 
 
 Xi 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -5 27 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 -4|28 
 
 
 
 
 
 4 
 
 
 
 
 
 
 
 
 
 4 
 
 
 
 -3 29 
 
 
 
 
 1 
 
 1 
 
 
 
 
 
 
 
 
 
 o 
 
 
 
 -2! 30 
 
 
 
 
 1 
 
 i 
 
 S 
 
 1 
 
 
 
 
 
 
 
 17 
 
 
 
 -1 31 
 
 
 
 
 3 
 
 11 
 
 3 
 
 5 
 
 1 
 
 
 
 
 
 
 09 1 
 
 
 
 
 32 
 
 
 
 
 2 
 
 7 
 
 14 
 
 14 
 
 4 
 
 
 
 
 
 
 41 
 
 
 1 
 
 33 
 
 
 
 
 1 
 
 2 
 
 6 
 
 8 
 
 17 
 
 o 
 
 1 
 
 
 
 
 37 
 
 
 2 34 
 
 
 
 
 
 3 
 
 2 
 
 2 
 
 3 
 
 2 
 
 
 
 
 
 12 
 
 
 3 35 
 
 
 
 
 
 
 3 
 
 4 
 
 2 
 
 
 1 
 
 
 
 
 10 
 
 
 4 ]36 
 
 
 
 
 
 1 
 
 
 
 1 
 
 3 
 
 
 
 
 
 5 
 
 
 5 j37 
 
 
 
 
 
 
 
 
 1 
 
 2 
 
 1 
 
 
 
 
 4 
 
 
 
 6 
 
 38 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 1 
 
 
 2 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 1 
 
 
 
 8 
 
 36 
 
 36 
 
 34 
 
 30 
 
 9 
 
 3 
 
 1 158 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Fig. 31. — Correlation in the buds oi Oenothera 
 lamarckiana. Xo. 04129. Thickness of ovary 
 subject, thickness of bud-cone relative. 
 Coefficient of correlation, 0.6506 ± 0.0309. 
 
MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 43 
 
 
 
 
 Xo 
 
 L4 
 
 ■: 
 
 -: 
 
 -1 
 
 
 
 1 
 
 
 
 3 
 
 
 
 
 
 
 
 
 05 
 
 70 
 
 7'> 
 
 80 
 
 85 
 
 90 
 
 95 
 
 100 
 
 
 
 
 
 
 
 
 TO 
 
 75 
 
 8U 
 
 85 
 
 00 
 
 95 
 
 100 
 
 105 
 
 
 
 
 
 X: 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -3 
 
 2S 
 
 
 
 
 
 
 4 
 
 
 1 
 
 
 
 
 7 
 
 
 
 — 
 
 2y 
 
 
 
 
 
 3 
 
 G 
 
 5 
 
 
 
 
 
 IG 
 
 
 
 -1 
 
 30 
 
 
 
 
 1 
 
 G 
 
 12 
 
 10 
 
 I 
 
 
 
 3G 
 
 
 
 
 
 31 
 
 
 1 
 
 
 
 14 
 
 10 
 
 i 
 
 
 
 
 
 34 
 
 
 
 1 
 
 32 
 
 
 
 
 2 
 
 2 
 
 12 
 
 7 
 
 / 
 
 
 
 30 
 
 
 
 
 
 33 
 
 
 
 
 
 
 S 
 
 4 
 
 5 
 
 
 
 
 00 
 
 
 
 3 
 
 3-1 
 
 
 
 
 
 
 
 
 
 G 
 
 1 
 
 
 10 
 
 
 
 4 
 
 35 
 
 
 
 
 
 
 1 
 
 
 
 
 I 
 
 
 4 
 
 
 
 5 
 
 30 
 
 
 
 
 
 
 
 
 1 
 
 
 
 1 
 
 
 
 (i 
 
 37 
 
 
 
 
 
 
 
 1 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 •1 
 
 14 
 
 ■il 
 
 4S 
 
 38 
 
 00 
 
 4 
 
 IGl 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Xo 
 
 -4 
 
 -3 
 
 _2 
 
 -1 
 
 
 
 I 
 
 2 
 
 3 4 5 
 
 G 
 
 
 ^^ 
 
 
 
 
 
 
 60 
 
 05 
 
 70 
 
 75 
 
 SO 
 
 85 
 
 90 
 
 95 
 
 100 
 
 105 
 
 110 
 
 
 
 
 
 
 
 
 05 
 
 70 
 
 75 
 
 80 
 
 85 
 
 90 
 
 95 
 
 100 
 
 105 
 
 110 
 
 115 
 
 
 
 
 
 X. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 _2 
 
 23 
 
 
 1 
 
 
 2 
 
 y 
 
 2 
 
 t> 
 
 
 
 
 
 
 
 16 
 
 
 
 -1 
 
 24 
 
 
 
 
 2 
 
 9 
 
 5 
 
 5 
 
 3 
 
 
 
 
 
 
 24 
 
 
 
 
 
 25 
 
 
 
 
 3 
 
 10 
 
 15 
 
 9 
 
 10 
 
 1 
 
 
 
 
 
 4S 
 
 
 
 1 
 
 2G 
 
 
 
 
 1 
 
 G 
 
 11 
 
 15 
 
 9 
 
 4 
 
 1 
 
 
 
 
 47 
 
 
 
 
 
 27 
 
 
 
 
 
 1 
 
 
 
 3 
 
 G 
 
 4 
 
 1 
 
 
 
 
 17 
 
 
 
 3 
 
 28 
 
 
 
 
 
 
 1 
 
 
 1 
 
 
 
 
 1 
 
 
 3 
 
 
 
 4 
 
 •29 
 
 
 
 
 
 1 
 
 
 
 1 
 
 
 1 
 
 
 
 
 3 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 8 
 
 3G 
 
 3G 
 
 34 
 
 30 
 
 9 
 
 3 
 
 
 
 1 
 
 
 158 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Fig. 32. — Correlation in buds of O. 
 lamarckiaiia. No. 04113. Thickness 
 of ovary subject, thickness of bud- 
 cone relative. Coefficient of cor- 
 relation. 0.5910 ±0.0346. 
 
 Fig. 34. — Correlation in the buds of Oenoiliera 
 lamarckia.na, No. 04129. Thickness of hy- 
 panthium subject, thickness of bud-cone rela- 
 tive. Coefficient of correlation, 0.5180 
 ± 0.0393. 
 
 
 
 
 Xol-2 
 
 -1 
 
 
 
 1 
 
 
 
 3 
 
 4 
 
 5 
 
 
 
 
 
 
 
 
 65 
 
 70 
 
 75 
 
 80 
 
 85 
 
 90 
 
 95 
 
 100 
 
 
 
 
 
 
 
 70 
 
 75 
 
 80 
 
 80 
 
 90 
 
 95 
 
 100 
 
 105 
 
 
 
 
 
 Xi 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -4 
 
 22 
 
 
 1 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 -3 
 
 23 
 
 
 1 
 
 <> 
 
 
 3 
 
 
 
 
 
 
 G 
 
 
 
 -2 
 
 24 
 
 
 3 
 
 4 
 
 
 
 1 
 
 1 
 
 
 
 
 
 11 
 
 
 
 -1 
 
 25 
 
 
 
 3 
 
 9 
 
 12 
 
 3 
 
 
 
 
 
 27 
 
 
 
 
 
 26 
 
 
 
 1 17 
 
 IG 
 
 10 
 
 3 
 
 
 
 
 
 49 
 
 
 
 1 
 
 27 
 
 
 
 
 10 
 
 15 
 
 17 
 
 5 
 
 
 
 
 47 
 
 
 
 2 
 
 28 
 
 
 
 
 
 6 
 
 7 
 
 2 
 
 
 1 
 
 
 IG 
 
 
 
 3 
 
 29 
 
 
 
 
 
 1 
 
 
 
 
 
 
 1 
 
 
 
 i 
 
 30 
 
 
 
 
 
 1 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 5 
 
 10 
 
 39 
 
 55 38 
 
 10 
 
 
 
 1 
 
 
 160 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 y^ 
 
 -4 
 
 -3 
 
 _o 
 
 -1 
 
 
 
 1 
 
 2 
 
 3 
 
 
 
 
 
 
 
 
 65 
 
 70 
 
 75 
 
 80 
 
 85 
 
 90 
 
 95 
 
 100 
 
 
 
 
 
 
 
 
 70 
 
 75 
 
 SO 
 
 85 
 
 90 
 
 95 100 
 
 105 
 
 
 
 
 
 X: 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 _3 
 
 00 
 
 
 
 1 
 
 1 
 
 
 
 
 
 
 
 •-> 
 
 
 
 _o 
 
 23 
 
 
 
 
 5 
 
 G 
 
 1 
 
 
 
 
 
 12 
 
 
 
 -1 
 
 24 
 
 
 
 1 
 
 4 
 
 7 
 
 i 
 
 4 
 
 
 
 
 23 
 
 
 
 
 
 25 
 
 
 1 
 
 
 2 
 
 9 
 
 17 
 
 13 
 
 / 
 
 
 
 49 
 
 
 
 1 
 
 2G 
 
 
 
 
 1 
 
 7 
 
 10 
 
 14 
 
 7 
 
 1 
 
 
 40 
 
 
 
 2 
 
 27 
 
 
 
 
 1 
 
 3 
 
 9 
 
 5 
 
 3 
 
 1 
 
 
 22 
 
 
 
 3 
 
 28 
 
 
 
 
 
 
 2 
 
 2 
 
 5 
 
 
 
 
 11 
 
 
 
 4 
 
 29 
 
 
 
 
 
 
 1 
 
 
 
 
 
 1 
 
 
 
 •1 
 
 30 
 
 
 
 
 
 
 1 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 I 
 
 2 
 
 14 
 
 32 
 
 48 38 
 
 22 
 
 4 
 
 
 IGl 
 
 
 
 
 
 
 
 
 
 
 . 
 
 
 
 
 Fig. 33. — Correlation in buds of O. lam- 
 arckiana from Hilversum. Thick- 
 ness of hypanthium subject, thick- 
 ness of bud -cone relative. Coefficient 
 of correlation, 0.5316 ± 0.0382. 
 
 Fig. 35. — Correlation in the buds of O. 
 lamarckiaiia. No. 04113. Thickness 
 of hypanthium subject, thickness 
 of bud-cone relative. Coefficient 
 of correlation, 0.5005 ±0.0398. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Xo 
 
 -9 
 
 -8 
 
 -7 
 
 -6 
 
 -5 
 
 -4 
 
 -3 
 
 -2 
 
 -1 
 
 
 
 1 
 
 2 
 
 3 
 
 4 
 
 5 
 
 6 
 
 7 
 
 8 
 
 9 
 
 
 
 
 
 
 
 
 21 i 22 
 
 23 
 
 24 
 
 25 
 
 26 
 
 27 '28 29 
 
 30 
 
 31 
 
 32 
 
 33 
 
 34 
 
 35 
 
 36 
 
 37 
 
 38 
 
 39 
 
 
 
 
 X. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -4 
 
 85 
 
 90 
 
 
 
 1 
 
 
 
 
 
 2 
 
 1 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 5 
 
 
 -3 
 
 90 
 
 95 
 
 
 
 
 2 
 
 
 
 1 
 
 3 
 
 
 2 
 
 
 
 2 
 
 
 
 
 
 1 
 
 
 
 
 11 
 
 
 _o 
 
 95 
 
 100 
 
 
 
 1 
 
 1 
 
 2 
 
 1 
 
 1 
 
 1 
 
 1 
 
 1 
 
 2 
 
 1 
 
 2 
 
 
 2 
 
 
 
 
 
 
 
 16 
 
 
 -1 
 
 1001C5 
 
 
 
 
 
 1 
 
 2 
 
 4 
 
 1 
 
 4 
 
 ty 
 
 4 
 
 8 
 
 2 
 
 1 
 
 2 
 
 
 1 
 
 1 
 
 
 
 
 33 
 
 
 
 
 105110 
 
 
 
 
 
 
 1 
 
 
 5 
 
 2 
 
 3 
 
 1 
 
 3 
 
 3 
 
 3 
 
 1 
 
 2 
 
 
 1 
 
 1 
 
 1 
 
 
 27 
 
 
 1 110115 
 
 
 1 
 
 
 
 
 2 
 
 1 
 
 1 
 
 2 
 
 1 
 
 5 
 
 3 
 
 
 1 
 
 2 
 
 2 
 
 4 
 
 3 
 
 1 
 
 1 
 
 
 30 
 
 
 2 ;1I5120 
 
 
 1 
 
 
 
 
 1 
 
 1 
 
 
 
 2 
 
 1 
 
 1 
 
 4 
 
 1 
 
 1 
 
 2 
 
 1 
 
 
 
 
 
 16 
 
 
 3 120125 
 
 
 
 
 
 1 
 
 
 
 
 
 1 
 
 1 
 
 2 
 
 3 
 
 1 
 
 1 
 
 1 
 
 
 
 
 
 
 11 
 
 
 4 
 
 125,130 
 
 
 
 
 
 
 
 
 
 
 1 
 
 1 
 
 1 
 
 2 
 
 
 3 
 
 
 1 
 
 
 
 
 
 9 
 
 
 5 
 
 130135 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 6 
 
 135 
 
 liO 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 2 
 
 2 
 
 3 
 
 4 
 
 8 
 
 8 
 
 13 
 
 10 
 
 13 
 
 15 
 
 20 
 
 18 
 
 7 
 
 13 
 
 7 
 
 7 
 
 6 
 
 2 
 
 
 
 IGO 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Fig. 36. — Correlation in the buds of Oenothera rubrinervis. No. 04113. 
 Length of ovary subject, length of hypanthium relative. Co- 
 efficient of correlation, . 2985 ± 0.0486. 
 
44 MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 
 
 
 ' 
 
 
 
 
 
 ~~ 
 
 ~ 
 
 
 
 ~ 
 
 
 
 
 
 
 
 
 
 
 
 
 n 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Xi 
 
 -I'J 
 
 -IS 
 
 -17 
 
 -IG 
 
 -15 
 
 -u 
 
 -13 
 
 -12 
 
 -11 
 
 -10 
 
 -9 
 
 -8 
 
 — i 
 
 -6 
 
 -5 
 
 -4 
 
 -3 
 
 o 
 
 -1 
 
 
 
 1 
 
 O 
 
 3 
 
 4 
 
 5 6 
 
 ■J" 
 
 
 
 
 
 
 
 
 11 
 
 12 
 
 13 
 
 14 
 
 15 
 
 16 
 
 17 
 
 18 
 
 19 
 
 20 
 
 21 
 
 22 
 
 23 
 
 24 
 
 25 
 
 26 
 
 27 
 
 28 
 
 29 
 
 30 
 
 31 
 
 32 
 
 33 
 
 34 
 
 35 "si? 
 
 37 
 
 
 
 
 X, 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 —5 
 
 90 
 
 95 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 -4 
 
 95 
 
 100 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 1 
 
 
 -3 
 
 100 
 
 105 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 1 
 
 3 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 6 
 
 
 -■Z 
 
 105 
 
 110 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 2 
 
 2 
 
 
 1 
 
 1 
 
 1 
 
 1 
 
 1 
 
 
 
 
 10 
 
 
 -1 
 
 110 
 
 115 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 4 
 
 4 
 
 1 
 
 o 
 
 O 
 
 4 
 
 o 
 
 O 
 
 1 
 
 
 
 
 
 23 
 
 
 
 
 115 
 
 120 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 1 
 
 ■J 
 
 5 
 
 4 
 
 4 
 
 - 
 
 ~ 
 
 O 
 
 1 
 
 4 
 
 
 1 
 
 
 35 
 
 
 1 
 
 120 
 
 125 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 4 
 
 3 
 
 2 
 
 4 
 
 6 
 
 O 
 
 (> 
 
 3 
 
 o 
 
 1 
 
 
 
 bl 
 
 
 o 
 
 125 
 
 130 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 1 
 
 1 
 
 3 
 
 3 
 
 3 
 
 4 
 
 3 
 
 
 
 
 
 20 
 
 
 3 
 
 13U 
 
 135 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 3 
 
 1 
 
 
 1 
 
 2 
 
 o 
 
 3 
 
 
 
 
 
 12 
 
 
 ■1 
 
 135 
 
 110 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 1 
 
 
 5 
 
 110 
 
 115 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 1 
 
 
 1 
 
 
 
 
 
 
 'J 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 ^ 
 
 
 
 1 
 
 2 
 
 7 
 
 15 
 
 15 
 
 13 14 , 19 
 
 IS 
 
 19 121 7 
 
 1 
 
 1 
 
 
 14G 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Fig. 37. — Correlation in the buds of Oenothera ruhrincrvis, Xo.'04137. Length of ovary subject, length 
 of hypanthium relative. Coefficient of correlation, 0.2211 ± 0.0531. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 i 
 
 
 
 
 
 
 
 
 
 
 
 Xc 
 
 -9 
 
 -S 
 
 -( 
 
 -c 
 
 -5 
 
 -4 
 
 -3 
 
 _o 
 
 -1 
 
 
 
 1 
 
 •"> 
 
 3 
 
 4 
 
 5 
 
 G 
 
 ( 
 
 S 
 
 9 
 
 10 
 
 11 
 
 12 
 
 
 
 
 
 
 
 
 26 
 
 27 
 
 28 
 
 29 
 
 30 
 
 31 
 
 32 
 
 33 
 
 34 
 
 35 
 
 36 
 
 37 
 
 38 
 
 39 
 
 40 
 
 41 
 
 42 
 
 43 
 
 44 
 
 45 
 
 46 
 
 47 
 
 
 
 
 X: 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -4 
 
 85 
 
 90 
 
 
 
 
 
 
 
 2 
 
 2 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 5 
 
 
 -3 
 
 90 
 
 95 
 
 
 
 
 1 
 
 1 
 
 1 
 
 1 
 
 
 2 
 
 2 
 
 2 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 11 
 
 
 _■■) 
 
 95 
 
 100 
 
 
 1 
 
 
 1 
 
 1 
 
 2 
 
 *> 
 
 o 
 
 2 
 
 1 
 
 3 
 
 
 •l 
 
 
 
 
 
 
 
 
 
 
 
 
 10 
 
 
 -1 
 
 100 
 
 105 
 
 
 
 1 
 
 
 
 
 
 2 
 
 3 
 
 5 
 
 5 
 
 i 
 
 5 
 
 3 
 
 2 
 
 
 
 
 
 
 
 
 
 
 33 
 
 
 
 
 105 
 
 110 
 
 
 
 
 
 1 
 
 1 
 
 1 
 
 1 
 
 1 
 
 5 
 
 5 
 
 4 
 
 3 
 
 3 
 
 
 
 2 
 
 
 
 
 
 
 
 
 27 
 
 
 1 
 
 110 
 
 115 
 
 
 
 
 
 
 
 1 
 
 1 
 
 1 
 
 3 
 
 O 
 
 6 
 
 
 
 4 
 
 2 
 
 4 
 
 1 
 
 
 
 
 
 
 
 
 30 
 
 
 •> 
 
 Ho 
 
 120 
 
 
 
 
 
 
 
 
 1 
 
 2 
 
 
 3 
 
 4 
 
 1 
 
 O 
 
 1 
 
 
 1 
 
 
 1 
 
 
 
 
 
 
 IG 
 
 
 3 
 
 120 
 
 125 
 
 
 
 
 
 
 
 
 
 
 
 
 O 
 
 2 
 
 5 
 
 1 
 
 1 
 
 
 
 
 
 
 
 
 
 11 
 
 
 4 
 
 125 
 
 130 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 1 
 
 
 2 
 
 1 
 
 
 o 
 
 1 
 
 
 
 
 1 
 
 
 9 
 
 
 5 
 
 130 
 
 135 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 6 
 
 135 
 
 110 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 160 
 
 
 
 
 
 
 1 
 
 1 
 
 O 
 
 3 
 
 4 
 
 i 
 
 9 
 
 12 
 
 16 
 
 21 
 
 24 
 
 19 
 
 17 
 
 8 
 
 G 
 
 5 
 
 2 
 
 o 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Fig. 38. — Correlation in the buds of Oenothera rubrinervis, Xo. 04113. Length of ovary subject, length 
 of bud-cone relative. Coefficient of correlation, 0.5976 ± 0.0343. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ~ 
 
 
 
 ~ 
 
 
 
 "~ 
 
 
 "-^ 
 
 
 
 
 
 X. 
 
 -11 
 
 40 
 
 -9 
 
 -8 
 
 -7 
 
 -6 
 
 -5 
 
 -4 
 
 -3 
 
 _o 
 
 -1 
 
 
 
 1 
 
 2 
 
 3 
 
 4 
 
 5 
 
 C 
 
 
 
 
 
 
 
 
 25 
 
 26 
 
 27 
 
 28 
 
 29 
 
 30 
 
 31 
 
 32 
 
 33 
 
 31 
 
 35 
 
 3G 
 
 37 
 
 3s 
 
 39 
 
 40 
 
 41 
 
 42 
 
 
 
 
 Xi 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -5 
 
 90 
 
 95 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 -4 
 
 95 
 
 100 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 -3 
 
 100 
 
 105 
 
 
 
 
 
 
 
 
 
 o 
 
 
 2 
 
 
 
 
 
 1 
 
 
 
 
 
 G 
 
 
 -2 
 
 105 
 
 no 
 
 
 1 
 
 
 
 
 
 
 
 
 
 2 
 
 
 o 
 
 o 
 
 1 
 
 
 1 
 
 
 
 
 HI 
 
 
 -1 
 
 110 
 
 115 
 
 
 
 
 
 
 
 
 
 1 
 
 
 5 
 
 
 5 
 
 4 
 
 3 
 
 1 
 
 1 
 
 
 
 
 23 
 
 
 
 
 115 
 
 120 
 
 
 
 
 
 
 
 
 
 
 
 3 
 
 
 6 
 
 7 
 
 G 
 
 3 
 
 1 
 
 1 
 
 1 
 
 
 35 
 
 
 1 
 
 120 
 
 125 
 
 
 
 
 
 
 
 
 
 
 1 
 
 5 
 
 
 10 
 
 O 
 
 5 
 
 4 
 
 2 
 
 1 
 
 
 
 34! 1 
 
 o 
 
 125130 
 
 
 
 
 
 
 
 
 
 
 1 
 
 3 
 
 
 1 
 
 4 
 
 3 
 
 2 
 
 1 
 
 
 1 
 
 
 20 
 
 
 3 
 
 130 
 
 133 
 
 
 
 
 
 
 
 
 
 
 1 
 
 O 
 
 
 o 
 
 3 
 
 O 
 
 
 
 1 
 
 
 
 12 
 
 4 
 
 135 
 
 110 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 1 
 
 
 5 
 
 UO 
 
 145 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 1 
 
 
 1 
 
 
 
 
 3 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 4 
 
 3 
 
 22 
 
 OO 
 
 27 
 
 23 
 
 21 
 
 U 
 
 4 
 
 3 
 
 2 
 
 
 146 
 
 
 
 _ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Fig. 39. — Correlation in the buds of Oenothera rubrinenns. No. 04137. Length of ovary subject, length 
 of bud-cone relative. Coefficient of correlation, 0.2464 ± 0.0524. 
 
MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 45 
 
 
 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 *""" 
 
 
 
 
 
 
 
 
 — 1 
 
 ~~ 
 
 
 "~ 
 
 
 ~ 
 
 
 
 ' 
 
 
 
 
 
 
 Xo 
 
 -4 
 
 -3 
 
 -2 
 
 -1 
 
 
 
 1 
 
 2 
 
 3 
 
 4 
 
 5 
 
 6 
 
 7 
 
 8 
 
 
 
 
 
 
 
 X. 
 
 -4 
 
 -3 
 
 -2 
 
 -1 
 
 
 
 1 
 
 2 
 
 3 
 
 4 
 
 5 
 
 G 
 
 / 
 
 
 
 
 
 
 
 
 29 
 
 30 
 
 31 
 
 32 
 
 33 
 
 34 
 
 35 
 
 36 
 
 37 
 
 38 
 
 39 
 
 40 
 
 41 
 
 
 
 
 
 
 
 
 32 
 
 33 
 
 34 
 
 35 
 
 36 
 
 37 
 
 38 
 
 39 
 
 40 
 
 41 
 
 42 
 
 43 
 
 
 
 
 Xi 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Xi 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -4185 
 
 90 
 
 
 1 
 
 
 
 1 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 5 
 
 
 -5 
 
 90 
 
 95 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 -3 
 
 90 
 
 95 
 
 
 1 
 
 1 
 
 3 
 
 4 
 
 
 1 
 
 
 
 
 1 
 
 
 
 
 
 U 
 
 
 -4 
 
 95 
 
 100 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 _o 
 
 95 
 
 100 
 
 
 2 
 
 
 
 1 
 
 7 
 
 2 
 
 ,1 
 
 
 1 
 
 
 
 
 
 
 
 16 
 
 
 -3 
 
 100 
 
 105 
 
 
 1 
 
 1 
 
 2 
 
 
 1 
 
 1 
 
 
 
 
 
 
 
 
 6 
 
 
 -1 
 
 100105 
 
 
 
 2 
 
 5 
 
 6 
 
 6 
 
 2 
 
 6 
 
 4 
 
 2 
 
 
 
 
 
 
 33 
 
 
 _o 
 
 105110 
 
 
 
 *> 
 
 2 
 
 2 
 
 
 
 
 
 1 
 
 1 
 
 
 
 
 
 10 
 
 
 
 
 105110 
 
 
 
 
 
 1 
 
 11 
 
 3 
 
 2 
 
 3 
 
 
 
 3 
 
 
 
 
 
 
 27 
 
 
 -I1UOII5 
 
 
 1 
 
 n 
 
 
 7 
 
 4 
 
 3 
 
 
 
 
 
 1 
 
 1 
 
 
 
 
 23 
 
 
 1 
 
 110115 
 
 
 1 
 
 
 1 
 
 4 
 
 7 
 
 8 
 
 3 
 
 2 
 
 3 
 
 1 
 
 
 
 
 
 30 
 
 
 115120 
 
 
 
 3 
 
 2 
 
 5 
 
 9 
 
 10 
 
 1 
 
 1 
 
 2 
 
 2 
 
 
 
 
 35 
 
 
 2 
 
 115i20 
 
 
 
 
 
 1 
 
 3 
 
 2 
 
 5 
 
 1 
 
 1 
 
 3 
 
 
 
 
 
 16 
 
 
 1 
 
 120jl25 
 
 
 
 
 1 
 
 G 
 
 8 
 
 10 
 
 
 
 1 
 
 1 
 
 
 1 
 
 
 
 34 
 
 
 3 
 
 120125 
 
 
 
 
 
 
 
 3 
 
 1 
 
 1 
 
 1 
 
 2 
 
 
 2 
 
 1 
 
 
 11 
 
 
 
 
 12.5130 
 
 
 
 
 
 4 
 
 9 
 
 3 
 
 1 
 
 1 
 
 1 
 
 1 
 
 
 
 
 20 
 
 
 4 
 
 125130 
 
 
 
 
 
 
 2 
 
 
 1 
 
 
 2 
 
 1 
 
 2 
 
 1 
 
 
 
 9 
 
 
 3 
 
 130135 
 
 
 
 
 
 
 
 
 
 4 
 
 1 
 
 
 1 
 
 1 
 
 
 1 
 
 
 12 
 
 
 5 
 
 13013.5 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 1 
 
 
 4 
 
 13.5 1110 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 1 
 
 
 6 
 
 135 
 
 UO 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 1 
 
 
 5 
 
 110 
 
 115 
 
 
 
 
 
 
 1 
 
 
 
 *> 
 
 
 
 
 
 
 3 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 5 
 
 9 
 
 12 
 
 34 
 
 23 
 
 19 
 
 19 
 
 12 13 8 
 
 2 
 
 3 1 
 
 IGO 
 
 
 
 
 
 
 
 
 10 
 
 7 
 
 26 
 
 34 
 
 32 
 
 13 
 
 8 
 
 i 
 
 5 
 
 1 
 
 1 
 
 
 146 
 
 
 
 
 
 
 
 
 
 
 
 
 
 i 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Fig. 40. — Correlation in the buds of Oenothera ru- 
 brinenis, No. 04113. Length of ovary subject, 
 thickness of ovary relative. Coefficient of correla- 
 tion, . 5957 ± . 0344. 
 
 Fig. 41. — Correlation in the buds of O. ruhrinervis , 
 Xo. 04137. Length of ovary subject, thickness 
 of ovary relative. Coefficient of correlation, 
 0.3357 ±0.0495. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Xo 
 
 -9 
 
 -8 
 
 i 
 
 -6 
 
 -6 
 
 -4 
 
 -3 
 
 _o 
 
 -1 
 
 
 
 1 
 
 2 
 
 3 
 
 4 
 
 5 
 
 6 7 
 
 8 
 
 9 
 
 10 
 
 U 12 
 
 
 
 
 
 
 
 
 26 
 
 27 
 
 28 
 
 29 
 
 30 
 
 31 
 
 32 
 
 33 
 
 34 
 
 35 
 
 36 
 
 37 
 
 38 
 
 39 
 
 40 
 
 41 
 
 42 
 
 43 
 
 44 
 
 45 
 
 46 
 
 47 
 
 
 
 
 
 X, 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -9 
 
 21 
 
 
 
 
 
 
 
 
 
 2 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 2 
 
 
 
 -8 
 
 22 
 
 
 
 
 
 
 1 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 2 
 
 
 
 -7 
 
 23 
 
 
 
 
 
 
 
 2 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 3 
 
 
 
 -6 
 
 24 
 
 
 
 
 
 
 
 1 
 
 1 
 
 
 
 
 1 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 4 
 
 
 
 -5 
 
 25 
 
 
 
 
 
 1 
 
 
 
 2 
 
 
 1 
 
 2 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 8 
 
 
 
 -4 
 
 26 
 
 
 
 
 
 1 
 
 
 
 
 
 3 
 
 1 
 
 2 
 
 
 
 
 
 
 
 
 
 
 
 
 
 8 
 
 — 
 
 
 -3 
 
 27 
 
 
 
 
 1 
 
 1 
 
 1 
 
 2 
 
 1 
 
 1 
 
 
 1 
 
 1 
 
 
 1 
 
 
 
 2 
 
 
 
 
 
 
 
 
 13 
 
 
 _o 
 
 28 
 
 
 1 
 
 1 
 
 
 
 
 1 
 
 
 
 1 
 
 3 
 
 1 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 10 
 
 
 
 -1 
 
 29 
 
 
 
 
 
 
 
 
 
 3 
 
 2 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 13 
 
 
 
 
 
 30 
 
 
 
 
 
 
 1 
 
 
 1 
 
 1 
 
 1 
 
 
 3 
 
 
 3 
 
 1 
 
 
 
 
 
 
 
 
 
 
 15 
 
 
 
 1 
 
 31 
 
 
 
 
 
 
 
 
 2 
 
 
 
 2 
 
 2 
 
 1 
 
 3 
 
 6 
 
 
 2 
 
 
 
 
 
 
 
 
 
 20 
 
 
 
 
 
 32 
 
 
 
 
 1 
 
 
 
 
 1 
 
 2 
 
 2 
 
 3 
 
 3 
 
 2 
 
 1 
 
 2 
 
 
 
 1 
 
 
 
 
 
 
 
 18 
 
 
 
 3 
 
 33 
 
 
 
 
 
 
 
 1 
 
 
 
 1 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 7 
 
 
 
 4 
 
 34 
 
 
 
 
 
 
 
 
 
 
 2 
 
 1 
 
 2 
 
 1 
 
 
 1 
 
 3 
 
 1 
 
 1 
 
 
 
 
 
 1 
 
 
 13 
 
 
 
 5 
 
 35 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 3 
 
 1 
 
 1 
 
 
 
 1 
 
 
 
 
 
 
 
 
 7 
 
 
 
 6 
 
 36 
 
 
 
 
 
 
 
 
 
 1 
 
 
 2 
 
 
 3 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 7 
 
 
 
 t 
 
 37 
 
 
 
 
 
 
 
 
 
 
 1 
 
 2 
 
 1 
 
 
 
 1 
 
 1 
 
 
 
 
 
 
 
 
 
 6 
 
 
 
 T 
 
 38 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 9 
 
 39 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 *> 
 
 
 
 
 
 
 
 
 
 
 
 
 ~ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 1 
 
 2 
 
 3 
 
 4 
 
 t 
 
 9 
 
 12 
 
 16 
 
 21 1 24 19 
 
 17 
 
 8 
 
 6 
 
 5 
 
 2 
 
 2 
 
 
 
 
 
 
 
 1 
 
 
 160 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 _ 
 
 
 _ 
 
 
 
 Fig. 42. — Correlation in the buds of Oenothera rubrinervis.tio. 0-i\\3. Length 
 of hypanthium subject, length of bud-cone relative. Coefficient of correla- 
 tion, 0.3752 ±0.0458. 
 
46 MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 f 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 c-i 
 
 CI 
 
 CO 
 
 -T 
 
 00 
 
 00 
 
 CO 
 
 o 
 
 -n 
 
 i-Ti 
 
 s 
 
 -Jj 
 
 t^ 
 
 2 1 «- 1 '- 
 
 CD 
 
 M 
 
 M 
 
 
 o 
 
 CD 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 IC 
 
 Si 
 
 
 
 
 
 
 
 
 -- 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 - 
 
 
 
 T 
 
 13 
 
 
 -^ 
 
 
 
 
 -< 
 
 
 
 - 
 
 -H 
 
 - 
 
 CO 
 
 - 
 
 - 
 
 eo 
 
 -^ 
 
 
 
 
 
 
 3 
 
 
 
 eo 
 
 CO 
 
 
 
 
 
 -H 
 
 - 
 
 - 
 
 - 
 
 
 ei 
 
 ta 
 
 
 - 
 
 
 - 
 
 '^ 
 
 - 
 
 
 
 - 
 
 
 CO 
 
 
 
 ri 
 
 o 
 
 CO 
 
 
 
 
 
 '- 
 
 CI 
 
 CI 
 
 -- 
 
 CJ 
 
 - 
 
 CO 
 
 - 
 
 CO 
 
 
 CO 
 
 ^ 
 
 ^ 
 
 -- 
 
 
 -^ 
 
 
 ?, 
 
 
 
 ^ 
 
 s 
 
 
 
 
 - 
 
 
 - 
 
 
 ^ 
 
 - 
 
 CI 
 
 C-l 
 
 CO 
 
 C) 
 
 -- 
 
 
 
 
 CI 
 
 -- 
 
 
 
 t^ 
 
 
 
 O 
 
 ?5 
 
 
 - 
 
 
 
 
 - 
 
 CI 
 
 CI 
 
 CI 
 
 CO 
 
 .- 
 
 CD 
 
 CO 
 
 M 
 
 11 
 
 eo 
 
 
 
 -- 
 
 
 
 CD 
 CO 
 
 
 
 1 
 
 
 
 
 CI 
 
 -^ 
 
 CJ 
 
 - 
 
 M 
 
 eo 
 
 c^l 
 
 CO 
 
 c^i 
 
 T 
 
 eo 
 
 C4 
 
 
 -■ 
 
 eo 
 
 
 
 
 
 C) 
 
 CO 
 
 
 
 CI 
 
 1 
 
 s 
 
 
 
 
 -H 
 
 
 
 
 CO 
 
 - 
 
 - 
 
 
 CO 
 
 C) 
 
 r^ 
 
 - 
 
 
 ^ 
 
 N 
 
 
 
 
 CO 
 
 
 
 CO 
 1 
 
 g? 
 
 
 
 
 
 
 - 
 
 
 - 
 
 
 
 
 
 CO 
 
 
 
 
 '- 
 
 •^ 
 
 
 
 
 '" 
 
 
 
 T 
 1 
 
 ?; 
 
 
 
 
 
 
 
 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 - 
 
 
 
 X 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Cl 
 
 ?i 
 
 ?5 
 
 C-1 
 
 s 
 
 25 
 
 s 
 
 r. 
 
 s 
 
 o 
 
 CO 
 
 CO 
 
 u 
 
 CO 
 CO 
 
 C5 
 
 S 
 
 CD 
 
 CO 
 
 00 
 
 CO 
 
 n 
 
 
 
 
 
 
 
 X 
 
 C5 
 
 CO 
 
 1 
 
 1^ 
 1 
 
 o 
 
 U3 
 
 1 
 
 T 
 
 CO 
 1 
 
 CI 
 
 1 
 
 .-H 
 1 
 
 o 
 
 -^ 
 
 CI 
 
 CO 
 
 ^ 
 
 »o 
 
 CD 
 
 l^ 
 
 00 
 
 C3 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 :3 
 
 c 
 
 P 
 
 cT 
 
 
 -^ 
 
 rt 
 
 n 
 
 
 
 
 •-.. 
 
 c, 
 
 
 
 
 c 
 
 >. 
 
 T' 
 
 a 
 
 
 
 
 
 
 
 
 
 
 
 
 c 
 
 
 n 
 
 u 
 
 
 c 
 
 G 
 
 
 '■J 
 
 
 r^ 
 
 j: 
 
 >i 
 
 
 
 
 
 
 
 3 
 
 c 
 
 o 
 
 
 
 
 
 
 JJ 
 
 h^ 
 
 rr. 
 
 ■*-* 
 
 
 O 
 
 
 o 
 
 Ci 
 
 
 j:: 
 
 
 c 
 
 a> 
 
 
 
 r^. 
 
 ^ 
 
 
 
 c 
 
 ^ 
 
 '-I 
 
 ia 
 
 
 
 
 
 
 
 u 
 
 o 
 
 
 
 
 •^ 
 
 ^, 
 
 
 - 
 
 i; 
 
 
 -^ 
 
 
 
 
 1^ 
 
 C3 
 
 ^ 
 
 o 
 
 
 o 
 
 
 
 
 
 
 u 
 
 ^ 
 
 n 
 
 
 
 u 
 
 
 
 ?-■ 
 
 ^ 
 
 > 
 
 +1 
 
 u 
 
 <:i> 
 
 
 .M 
 
 
 1 
 
 ;:: 
 
 p2 
 
 -*-> 
 
 on 
 
 
 r" 
 
 S 
 
 a 
 
 
 •+ 
 
 -c, 
 
 -^ 
 
 •—I 
 
 o 
 
 ■* 
 
 i; 
 
 ^ 
 
 o 
 
 
 
 ** 
 
 +-< 
 
 >-. 
 
 u 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ^ o 
 
 o 
 
 o 
 
 o 
 
 o 
 
 o 
 
 o 
 
 ■o 
 
 o 
 
 —1 
 
 o 
 
 o 
 
 - 
 
 CI 
 
 1^ 
 
 I/? 
 
 lO 
 
 CO 
 
 T 
 
 Cl 
 
 X 
 
 rs 
 
 M 
 
 t- 
 
 1-H 
 
 -^ 
 
 
 o 
 -r 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 :3 
 
 C) 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 rH 
 
 
 - 
 
 
 
 
 CI 
 
 
 
 
 O 
 
 •^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 tH 
 
 
 -^ 
 
 -^ 
 
 
 CO 
 
 
 
 
 ^ 
 
 c: 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 eo 
 
 CO 
 
 T-< 
 
 
 
 
 
 t- 
 
 
 
 
 CO 
 
 CO 
 
 
 ^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 CI 
 
 c^ 
 
 -r 
 
 c^ 
 
 
 
 
 -H 
 
 
 
 
 CI 
 
 CO 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 CI 
 
 T 
 
 co 
 
 CO 
 
 -J 
 
 d 
 
 r-t 
 
 
 
 
 c3 
 
 
 
 
 -H 
 
 s 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -T 
 
 - 
 
 M 
 
 CO 
 
 -r 
 
 c^ 
 
 - 
 
 - 
 
 
 
 
 ?3 
 
 
 
 
 o 
 
 CO 
 CO 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -^ 
 
 -- 
 
 ^ 
 
 CO 
 
 _ 
 
 CO 
 
 c> 
 
 ^ 
 
 CO 
 
 Cfl 
 
 C<l 
 
 « 
 
 
 
 
 s 
 
 
 
 
 T 
 
 CO 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 CO 
 
 CO 
 
 CO 
 
 CO 
 
 -> 
 
 C) 
 
 CO 
 
 
 -• 
 
 
 
 
 CI 
 CI 
 
 
 
 
 CI 
 
 1 
 
 CO 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1-1 
 
 CO 
 
 IC 
 
 CO 
 
 CO 
 
 N 
 
 eo 
 
 - 
 
 
 -■ 
 
 
 
 
 
 CJ 
 
 
 
 
 CO 
 
 1 
 
 CO 
 CO 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ^ 
 
 ^ 
 
 .-t 
 
 
 
 
 
 
 
 
 
 CO 
 
 
 
 
 I* 
 1 
 
 CI 
 
 CO 
 
 
 
 
 
 
 
 
 
 
 
 
 rH 
 
 
 
 
 
 •M 
 
 r-* 
 
 
 
 
 
 
 
 
 
 
 
 
 T 
 
 
 
 
 1 
 
 CO 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 o 
 o 
 
 
 
 
 CO 
 
 u 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 ci 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 o 
 
 
 
 
 CO 
 
 1 
 
 OO 
 CI 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 o 
 
 
 
 
 C5 
 
 s 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 o 
 
 
 
 
 o 
 
 s 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 o 
 
 
 
 
 ^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 rH 
 
 
 
 
 
 
 
 
 
 
 
 
 
 - 
 
 
 
 
 11 
 
 X 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ;::; 
 
 CI 
 
 1-i 
 
 CO 
 
 T 
 
 lo 
 
 CD 
 
 t* 
 
 CO 
 
 en 
 
 ?5 
 
 CI 
 
 ?! 
 
 CO 
 C4 
 
 s 
 
 s 
 
 s 
 
 S5 
 
 00 
 CI 
 
 CI 
 CI 
 
 s 
 
 CO 
 
 CI 
 
 CO 
 
 CO 
 CO 
 
 -r 
 CO 
 
 i? 
 
 CO 
 CO 
 
 
 
 
 
 
 
 
 
 >< 
 
 Ci 
 
 CO 
 
 r- 
 
 CD 
 
 i-H 
 
 
 TT 
 
 CO 
 
 M 
 
 
 o 
 
 Ci 
 
 I 
 
 00 
 
 1 
 
 1 
 
 CO 
 
 1 
 
 1 
 
 CO 
 
 1 
 
 CI 
 
 1 
 
 T 
 
 o 
 
 - 
 
 CI 
 
 CO 
 
 ■V 
 
 o 
 
 CO 
 
 t^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ' 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 O C^ 
 
 ? o 
 
 ^ o 
 
 
 u o 
 
 :5 
 
 3 -H 
 
 ^ 
 
 . c> 
 
 s^ 
 
 ■M O 
 
 o 
 
 
 C 3 o 
 
 .2 p e" 
 
 3 
 
 -1 
 
 
 
 
 
 
 '" 
 
 ^ 
 
 n! 
 
 
 
 
 o 
 
 C 
 
 
 ^ 
 
 rt 
 
 Ut 
 
 ■*-' 
 
 a 
 
 
 
 
 >. 
 
 U 
 
 C 
 
 
 
 •ii 
 
 t^ 
 
 
 
 c 
 
 
 
 ■*J 
 
 
 
 
 
 
 
 ■*-J 
 
 -M 
 
 o 
 
 
 
 (.J 
 
 
 3 
 
 s 
 
 
 
 1 
 
 r-.' 
 
 u 
 
 f^' 
 
 CO 
 
 0) 
 
 > 
 
 ■* 
 
 •c^ 
 
 
 o 
 
 -*-> 
 
 « 
 
 
 
 
 
 
 tn 
 
 
 
MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 47 
 
 
 
 
 
 '*; 
 
 
 
 
 
 
 
 r 
 
 ^^ 
 
 ^^ 
 
 ^ 
 
 
 — 
 
 — 
 
 
 4^ 
 
 
 
 
 
 
 P 
 
 C 
 
 i 
 
 ^ 
 
 
 
 3 
 
 3 
 
 5' 
 
 1 
 
 
 
 s 
 
 p 
 
 ■t 
 
 
 
 rr* 
 
 
 
 ^- 
 
 ^. 
 
 
 
 
 
 £■ 
 
 tn 
 
 < 
 
 
 4^ 
 
 K) 
 
 
 
 ^ ' 
 
 "^ 
 
 •T' 
 
 r\ 
 
 
 
 
 
 H- 
 
 
 
 
 n' 
 
 
 5" 
 
 
 
 fn 
 
 3 
 
 3 
 
 
 c 
 
 01 
 
 0' 
 
 'J5 
 
 r; 
 
 V; 
 
 " 
 
 3 
 
 ^ 
 
 *-4 
 
 n 
 
 
 
 
 
 
 "^ 
 
 
 c 
 
 
 
 
 
 
 
 
 
 
 
 
 *<: 
 
 v^ 
 
 ^ 
 
 
 
 
 
 
 
 
 ^ 
 
 w 
 
 ^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ■ 
 
 
 
 
 
 
 
 -^I 
 
 
 
 CT 
 
 >(;.. 
 
 M 
 
 to 
 
 - 
 
 
 
 1 
 
 ■ i ■ 
 
 to 1 CO 
 
 1 
 
 1 
 
 1 
 
 1 
 
 -J 
 
 1 
 
 00 
 
 1 
 
 CD 
 
 
 
 H- 
 
 
 
 ii 
 
 c: 
 
 
 cc 
 
 t3 
 
 x 
 
 
 
 
 
 
 
 £3 
 
 
 y 
 
 
 
 
 0; 
 
 Ci 
 
 
 
 1^ 
 
 cc 
 
 It 
 
 to 
 
 to 
 
 en 
 
 to 
 
 10 
 
 05 
 
 to 
 to 
 
 to 
 
 
 
 C2 
 
 OC 1 M 
 
 c: 
 
 j; 
 
 0^ 
 
 CO 
 
 to 
 
 ^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 X 
 
 
 
 CO 
 
 
 
 
 
 
 
 
 
 - 
 
 
 
 (-• 
 
 ^- 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ti 
 
 1 
 
 
 I ~ 
 
 
 
 
 15 
 
 »-- 
 
 
 
 - 
 
 
 
 
 - 
 
 
 
 
 
 
 p— 
 
 
 
 
 
 
 
 
 
 
 
 
 I; 
 
 1 
 
 
 
 1^ 
 
 
 
 
 
 
 a 
 
 •L. 
 
 u 
 
 
 w 
 
 w 
 
 to 
 
 w 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 to 
 
 ^ 
 
 
 
 w 
 
 
 
 - 
 
 tc 
 
 c 
 
 •I^ 
 
 Cl 
 
 en 
 
 cc 
 
 CT 
 
 ti- 
 
 ^1 
 
 tc 
 
 - 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 - 
 
 
 CO 
 
 
 
 
 
 
 
 0; 
 
 
 
 
 CO 
 
 "* 
 
 ^ c= 
 
 ^I 
 
 C-- ; t; 1 ^ ;; 1 - 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 CJ 
 
 - 
 
 
 
 s 
 
 
 
 
 
 - 
 
 >U 
 
 .£. 
 
 -- 
 
 >- " 
 
 CO 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 C-3 
 
 to 
 
 t-5 
 
 
 
 00 
 
 
 *-• 
 
 
 
 
 
 - 
 
 " 
 
 - 
 
 
 - 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 S! 
 
 W 
 
 
 
 - 
 
 
 
 
 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Ci 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 B 
 
 - 
 
 >- 
 
 " Hi £ 
 
 a:; 
 
 cc 
 
 ■C 
 
 ^ 
 
 0! 
 
 r; -t 
 
 i: 
 
 ^ 
 
 
 
 = 
 
 - 
 
 
 
 C 
 
 C 
 
 = 
 
 
 
 c 
 
 c c 
 
 
 
 - 
 
 
 
 
 
 
 
 ' 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 sub: 
 live 
 ±0 
 
 % 
 
 4^ 
 
 
 
 1 
 
 
 
 
 
 2?--^ 
 
 
 ( i 
 
 2; 
 
 
 
 
 
 n 
 
 3 5: 
 
 
 /^ 
 
 c r. 
 
 .r. 
 
 — 
 
 
 
 
 
 2 
 
 ^ 
 
 r 
 
 •-»» 
 
 
 
 •-1 
 
 
 3 
 
 
 
 ^ 
 
 IQ 
 
 
 
 
 f5" 
 
 
 
 2 
 
 s. 
 
 B 
 
 cr 
 
 
 
 
 c 
 
 
 
 
 c. 
 
 no 
 
 "-I 
 
 n 
 
 Oi 
 
 
 
 C 
 
 C/i 
 
 r." 
 
 
 Cm 
 
 1 
 
 c 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 IC 
 
 - 
 
 s 
 
 C£ 
 
 00 
 
 ^ 
 
 C5 
 
 c 
 
 J^ 
 
 CO 
 
 to 
 
 - 
 
 
 
 j^ 
 
 1 
 to 
 
 1 
 
 CO 
 
 1 
 
 1 
 
 c-1 
 
 ! 
 
 C3 
 
 1 
 
 cc 
 
 r 
 
 X 
 
 
 
 
 
 
 
 1^ 
 
 
 
 
 CO 
 
 to 
 
 i 
 
 
 CO 
 
 g 
 
 CO 
 *4 
 
 CO 
 
 g 
 
 CO 
 
 ss 
 
 CO 
 CO 
 
 OS 
 
 § 
 
 g 
 
 s 
 
 13 
 
 g 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 X 
 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 - 
 
 
 s 
 
 -I 
 
 
 
 >£- 
 
 
 01 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 - 
 
 
 
 
 t: 
 
 *-» 
 
 k- 
 
 
 
 00 
 
 
 
 S! 
 
 1 
 CO 
 
 15 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 l-» 
 
 CO 
 
 >&. 
 
 j^ 
 
 to 
 
 >-* 
 
 - 
 
 
 
 
 ■& 
 
 g 
 
 1 
 
 lis 
 
 
 
 
 
 
 
 
 
 
 
 
 - 
 
 CO 
 
 o- 
 
 c 
 
 Cl 
 
 CO 
 
 CO 
 
 I-* 
 
 
 
 
 
 
 
 
 JS 
 
 J_ 
 
 
 Z 
 
 
 
 
 
 
 
 
 
 
 
 0,1 
 
 h-t 
 
 •fe 
 
 CO 
 
 -vj 
 
 10 
 
 - 
 
 
 t-» 
 
 
 
 
 
 
 p 
 
 
 
 c 
 
 
 &3 
 
 
 
 
 
 
 
 
 - 
 
 - 
 
 CO 
 
 c; 
 
 CO 
 
 .£. 
 
 
 to 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 - 
 
 a 
 
 
 
 
 
 
 
 
 
 CO 
 
 CT 
 
 Cl 
 
 CO 
 
 ts 
 
 CO 
 
 
 
 
 
 
 
 
 
 
 
 8 
 
 8 
 
 t3 
 
 
 ts 
 
 
 I-* 
 
 
 
 
 H- 
 
 to 
 
 to 
 
 to 
 
 
 i-» 
 
 )-> 
 
 H' 
 
 
 
 
 ^- 
 
 
 
 
 
 
 
 
 
 
 g 
 
 C3 
 
 
 CO 
 
 
 
 
 
 
 1— 
 
 
 to 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Cn 
 
 
 
 .t- 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 § 
 
 
 »— 
 
 c 
 
 1= 
 
 =: 
 
 to 
 
 to 
 
 c 
 
 
 
 OD 
 
 r, 1 s i ^ 
 
 ro 
 
 c; 
 
 to 
 
 c: 
 
 ^I 
 
 ^ 
 
 CO 
 
 tci 
 
 - 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 ! 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ►^ 
 
 
 
 
 
 
 
 
 
 
 n 
 
 
 
 
 c 
 
 t*- 
 
 
 
 - 
 
 
 ■^ 
 
 *. 
 
 
 n 
 
 
 s 
 
 ^ 
 
 
 
 ;? 
 
 
 
 r. 
 
 nn 
 
 p 
 
 
 
 
 
 
 <■ 
 
 3 
 
 2. 
 
 
 
 
 
 
 
 
 
 H- 
 
 
 C 
 
 3 
 
 5-. 
 
 
 
 
 
 0* 
 
 
 2 
 
 b 
 
 s 
 
 
 25 
 
 -. 
 
 
 
 ^ 
 
 
 
 "^ 
 
 tn 
 
 r; 
 
 •-^ 
 
 
 '^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 .*^ 
 
 
 
 (— ■ 
 
 7f 
 
 
 _^ 
 
 
 
 "• 
 
 t« 
 
 
 
 
 
 
 
 ^ 
 
 
 
 
 w 
 
 
 
 
 r. 
 
 c 
 
 — 1 
 
 1-^ 
 V 
 
 ^ 
 
 2. 3 J5 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 a 
 
 CT 
 
 •c^ 
 
 CO 
 
 to 
 
 - 
 
 
 
 J^ 
 
 i 
 
 1 
 
 CO 
 
 1 
 
 4- 
 
 1 
 
 CT 
 
 d: 
 
 1 
 -I 
 
 1 
 or 
 
 cr 
 
 c 
 
 i 
 
 X 
 
 
 
 
 
 
 
 t3 
 
 .i- 
 
 
 
 g 
 
 §? 
 
 ^ 
 
 g 
 
 CO 
 
 CT 
 
 cc 
 
 CO 
 CO 
 
 co 
 £•5 
 
 £:!§ 
 
 
 '& 
 
 to 
 
 55 
 
 to 
 
 Cl 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 X 
 
 1^ 
 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 t— 
 
 
 -J 
 
 -.J 
 
 
 1 
 
 
 c 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 00 
 
 
 
 Cl 
 
 1 
 
 CO 
 
 
 to 
 
 
 
 
 
 
 
 
 
 
 l-t 
 
 
 »-» 
 
 
 
 
 
 
 
 
 
 CO 
 Cl 
 
 00 
 
 1 
 to 
 
 
 ij 
 
 
 
 
 
 
 
 
 CO OC 
 
 to 
 
 ta 
 
 to 
 
 
 
 
 
 
 
 
 
 § 
 
 00 
 
 1 
 
 
 OT 
 
 
 
 
 
 M 
 
 M 
 
 :r 
 
 s; 1 s 
 
 cc 
 
 K* 
 
 ^ 
 
 
 
 
 
 
 
 
 
 g 
 
 g 
 
 
 
 
 w 
 
 
 
 
 to 
 
 CT 
 
 CO 
 
 CO 
 
 00 
 
 to 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 g 
 
 ■^ 
 
 
 13 
 
 
 - 
 
 to 
 
 Cl 
 
 tU 
 
 CT 
 
 w 
 
 - 
 
 l-l 
 
 
 
 
 
 
 
 
 
 
 
 
 i 
 
 g 
 
 N) 
 
 
 to 
 
 
 ^ 
 
 t— 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 g 
 
 CO 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 55 
 
 
 t5 
 
 CO 
 
 <I 
 
 n: 
 
 to 
 
 to 
 
 10 
 -J 
 
 Ei 
 
 11 
 
 w 
 
 .£. 
 
 c 
 
 
 
 <= 
 
 
 
 c 
 
 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
48 MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 X, 
 
 -4 
 
 -3 
 
 _o 
 
 -1 
 
 
 
 1 
 
 O 
 
 o 
 
 4 
 
 5 
 
 
 
 
 
 
 
 
 24 
 
 25 
 
 2G 
 
 27 
 
 28 
 
 29 
 
 30 
 
 31 
 
 oo 
 
 33 
 
 
 
 
 
 X, 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -4 
 
 1\) 
 
 
 
 
 
 4 
 
 1 
 
 
 
 
 
 
 
 5 
 
 
 
 — J 
 
 30 
 
 
 1 
 
 3 
 
 1 
 
 4 
 
 
 
 
 
 
 
 
 9 
 
 
 
 _o 
 
 31 
 
 
 
 1 
 
 5 
 
 2 
 
 *> 
 
 2 
 
 
 
 
 
 
 12 
 
 
 
 -1 
 
 32 
 
 
 
 2 
 
 9 
 
 9 
 
 10 
 
 2 
 
 
 1 
 
 1 
 
 
 
 34 
 
 
 
 
 
 33 
 
 
 
 
 1 
 
 U 
 
 9 
 
 >> 
 
 3 
 
 2 
 
 
 
 
 23 
 
 
 
 1 
 
 34 
 
 
 
 
 
 6 
 
 G 
 
 4 
 
 2 
 
 1 
 
 1 
 
 
 
 19 
 
 
 
 2 
 
 3j 
 
 
 
 1 
 
 
 1 
 
 1 
 
 5 
 
 7 
 
 1 
 
 (J 
 
 
 
 19 
 
 
 
 o 
 
 3G 
 
 
 
 
 
 
 5 
 
 1 
 
 (i 
 
 
 
 
 
 12 
 
 
 
 4 
 
 37 
 
 
 
 
 
 
 o 
 
 1 
 
 4 
 
 3 
 
 O 
 
 1 
 
 
 13 
 
 
 
 5 
 
 38 
 
 
 
 
 
 1 
 
 
 
 
 2 
 
 5 
 
 
 
 * 
 
 
 
 G 
 
 3'J 
 
 
 
 
 
 
 
 
 
 1 
 
 1 
 
 
 
 
 
 
 i 
 
 40 
 
 
 
 
 
 
 
 
 1 
 
 1 
 
 1 
 
 
 
 3 
 
 
 
 8 
 
 41 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 7 
 
 IG 
 
 32 
 
 3G 
 
 IT 
 
 23 
 
 13 
 
 14 
 
 1 
 
 
 160 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Fig. 48. — Correlation in the buds of Oenothcrj, 
 rubrinenis. No. 04113. Thicknessof ovary 
 subject, thickness of hypanthium relative. 
 CoefiBcient of correlation, . 7026 ± . 0270.- 
 
 
 
 
 
 
 
 
 
 
 1 
 
 ^^ 
 
 
 
 
 
 x. 
 
 _3 
 
 -2 
 
 -1 
 
 
 
 1 
 
 O 
 
 3 
 
 4 
 
 
 
 
 
 
 
 
 27 
 
 28 
 
 29 
 
 30 
 
 31 
 
 32 
 
 33 
 
 34 
 
 
 
 
 
 X, 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -4 
 
 32 
 
 
 1 
 
 
 1 
 
 
 
 
 
 
 
 2 
 
 
 
 -3 
 
 33 
 
 
 1 
 
 O 
 
 5 
 
 '"> 
 
 
 
 
 
 
 10 
 
 
 
 __o 
 
 34 
 
 
 
 1 
 
 
 4 
 
 2 
 
 
 
 
 
 / 
 
 
 
 -1 
 
 35 
 
 
 1 
 
 1 
 
 5 
 
 9 
 
 i 
 
 2 
 
 1 
 
 
 
 2G 
 
 — 
 
 
 
 
 36 
 
 
 
 1 
 
 7 
 
 17 
 
 t 
 
 2 
 
 
 
 
 34 
 
 
 1 
 
 37 
 
 
 
 
 o 
 
 12 
 
 10 
 
 7 
 
 1 
 
 
 
 32 
 
 
 
 2 
 
 3S 
 
 
 
 
 3 
 
 5 
 
 2 
 
 3 
 
 
 
 
 13 
 
 
 
 3 
 
 39 
 
 
 
 
 
 2 
 
 3 
 
 
 2 
 
 1 
 
 
 8 
 
 
 
 4 
 
 40 
 
 
 
 1 
 
 
 1 
 
 3 
 
 
 o 
 
 
 
 t 
 
 
 
 5 
 
 41 
 
 
 
 
 
 1 
 
 1 
 
 1 
 
 n 
 
 
 
 5 
 
 — 
 
 
 6 
 
 42 
 
 
 
 
 
 
 
 1 
 
 
 
 
 1 
 
 
 7 
 
 43 
 
 
 
 
 
 
 1 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 3 
 
 6 
 
 23 63 
 
 3G 
 
 IG 
 
 8 
 
 1 
 
 
 146 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Fig. 49. — Correlation in the buds of 
 Oenothera rubrinervis, No. 04137. 
 Thickness of ovary subject, thick- 
 ness of hypanthium relative. Coef- 
 ficient of correlation, . 504 1 ± 0. 04 1 6. 
 
 
 
 
 Xo 
 
 -4 
 
 -3 
 
 _o 
 
 -1 
 
 
 
 1 
 
 o 
 
 3 
 
 4 
 
 ^~~ 
 
 ■~" 
 
 
 
 
 
 
 70 
 
 73 
 
 80 
 
 85 
 
 90 
 
 95 
 
 100 
 
 105 
 
 110 
 
 
 
 
 
 
 
 
 75 
 
 SO 
 
 85 
 
 <J0 
 
 05 
 
 100 
 
 105 
 
 110 
 
 113 
 
 
 
 
 X, 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -4 
 
 29 
 
 
 
 
 2 
 
 ."> 
 
 1 
 
 
 
 
 
 
 5 
 
 
 
 -3 
 
 30 
 
 
 1 
 
 1 
 
 4 
 
 o 
 
 1 
 
 
 
 
 
 
 9 
 
 
 
 — 2 
 
 31 
 
 
 
 
 4 
 
 5 
 
 1 
 
 1 
 
 1 
 
 
 
 
 12 
 
 
 
 -1 
 
 32 
 
 
 
 4 
 
 o 
 
 9 
 
 13 
 
 3 
 
 2 
 
 
 
 
 34 
 
 
 
 
 
 33 
 
 
 
 
 2 
 
 2 
 
 10 
 
 5 
 
 1 
 
 3 
 
 
 
 23 
 
 
 
 1 
 
 34 
 
 
 
 
 1 
 
 3 
 
 8 
 
 4 
 
 O 
 
 1 
 
 
 
 19 1 
 
 
 2 
 
 35 
 
 
 
 
 
 3 
 
 
 
 5 
 
 7 
 
 2 
 
 
 
 19 
 
 
 
 3 
 
 36 
 
 
 
 
 
 1 
 
 4 
 
 5 
 
 ■> 
 
 
 
 
 12 
 
 
 
 4 
 
 37 
 
 
 
 
 
 
 3 
 
 4 
 
 ■> 
 
 3 
 
 1 
 
 
 13 
 
 
 
 5 
 
 3S 
 
 
 
 
 
 
 1 
 
 2 
 
 .-) 
 
 1 
 
 2 
 
 
 8 
 
 
 
 6 
 
 39 
 
 
 
 
 
 
 
 
 
 2 
 
 
 
 2 1 
 
 
 7 
 
 40 
 
 
 
 
 
 
 
 1 
 
 2 
 
 
 
 
 O 
 
 
 
 8 
 
 41 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 5 
 
 16 
 
 27 
 
 44 
 
 31 
 
 21 
 
 12 
 
 3 
 
 
 16U 
 
 
 
 
 
 
 
 — 
 
 1 
 
 
 
 
 
 
 
 
 
 Fig. 50. — Correlation in the buds of Oeno- 
 thera rubrinervis, 'Ho. 04113. Thickness 
 of ovary subject, thickness of cone rel- 
 ative. Coefficient of correlation, 0.6310 
 ± 0.0321. 
 
 
 
 IXnl-4!-S 
 
 -2 
 
 -1 
 
 
 
 1I2U 
 
 
 
 
 
 
 
 
 70 
 
 75 
 
 SO 
 
 S5 
 
 90 
 
 95 10(^105 
 
 
 
 
 
 
 
 
 75 
 
 SO 
 
 S3 
 
 90 
 
 93|100 
 
 103 
 
 110 
 
 
 
 
 
 X, 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -4 32 
 
 
 
 
 1 
 
 
 1 
 
 
 
 
 
 
 
 
 
 — o <io 
 
 
 1 
 
 
 
 3 
 
 6 
 
 
 
 
 
 10 
 
 
 
 -2 31 
 
 
 
 
 
 O 
 
 3 
 
 1 
 
 1 
 
 
 
 1 
 
 
 
 -1 
 
 35 
 
 
 
 
 1 
 
 8 
 
 11 
 
 3 
 
 
 
 
 
 2G 
 
 
 
 
 
 36 
 
 
 
 
 
 10 
 
 17 
 
 7 
 
 
 
 
 34 
 
 
 
 1 
 
 37 
 
 
 
 
 
 4 
 
 7 
 
 12 
 
 8 
 
 1 
 
 
 32 
 
 
 
 2 
 
 38 
 
 
 
 
 
 
 3 
 
 7 
 
 •"> 
 
 1 
 
 
 13 
 
 
 
 3 
 
 39 
 
 
 
 
 
 
 
 5 
 
 3 
 
 
 
 8 
 
 
 
 4 
 
 40 
 
 
 
 
 
 
 •> 
 
 4 
 
 1 
 
 
 
 7 
 
 
 
 5 
 
 41 
 
 
 
 
 
 
 
 3 
 
 •y 
 
 
 
 5 
 
 
 
 6 
 
 42 
 
 
 
 
 
 
 1 
 
 
 
 
 
 1 
 
 
 
 7 
 
 43 
 
 
 
 
 
 
 
 1 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 2 
 
 27 
 
 51 
 
 43 
 
 20 
 
 '1 
 
 
 146 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Fio. 51. — Correlation in the buds of Oeno- 
 thera rubrinervis, Xo. 04137. Thick- 
 ness of ovary subject, thickness of 
 bud-cone relative. Coefficient of cor- 
 relation, 0.5006 ±0.0418. 
 
MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE) OENOTHERAS. 49 
 
 
 
 
 X2-4 
 
 -3 
 
 _o 
 
 -1 
 
 
 
 1 
 
 
 
 3 
 
 -> 
 
 
 
 
 
 
 
 
 70 
 
 75 
 
 80 
 
 85 
 
 9J 
 
 95 
 
 100 
 
 105 
 
 U« 
 
 
 
 
 
 
 
 
 75 
 
 80 
 
 85 
 
 90 
 
 95 
 
 100 
 
 105 
 
 110 
 
 115 
 
 
 
 
 
 Xi 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -4 
 
 24 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 -3 
 
 25 
 
 
 
 2 
 
 
 3 
 
 2 
 
 
 
 
 
 
 1 
 
 
 
 —2 
 
 26 
 
 
 
 1 
 
 5 
 
 5 
 
 3 
 
 1 
 
 1 
 
 
 
 
 IG 
 
 
 
 -127 
 
 
 
 1 
 
 10 
 
 8 
 
 12 
 
 1 
 
 
 
 
 
 32 
 
 
 
 |2S 
 
 
 
 1 
 
 1 
 
 i 
 
 10 
 
 8 
 
 
 
 1 
 
 
 
 3(i 
 
 
 
 1 
 
 21) 
 
 
 
 
 
 3 
 
 
 
 rt 1 
 
 1 
 
 
 
 17 1 
 
 
 2 
 
 30 
 
 
 
 
 
 
 5 
 
 9 
 
 G 
 
 3 
 
 
 
 23' 
 
 
 3 
 
 31 
 
 
 
 
 
 
 4 
 
 3 
 
 3 
 
 2 
 
 1 
 
 
 13 1 
 
 
 4 
 
 32 
 
 
 
 
 
 1 
 
 
 3 
 
 4 
 
 4 
 
 2 
 
 
 14 
 
 
 5 
 
 33 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 5 
 
 16 
 
 27 
 
 44 
 
 31 21 
 
 12 
 
 3 
 
 
 160 
 
 
 
 
 L_ 
 
 
 Mil! 
 
 
 
 
 i 
 
 1 
 
 
 
 
 X, 
 
 -4 
 
 -3-2I-I 
 
 
 
 l|2 
 
 3 
 
 
 
 
 
 
 
 
 70 
 
 75 
 
 SO ! 85 
 
 90 95 'lOoios 
 
 
 
 
 
 
 
 
 75 
 
 80 
 
 85190 
 
 9oliooio5ino 
 
 
 
 
 
 X, 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -3 
 
 27 
 
 
 1 
 
 
 2 
 
 
 
 
 
 
 
 3 
 
 
 
 -2 
 
 28 
 
 
 
 
 
 2 
 
 9 
 
 2 
 
 
 
 
 6 
 
 
 
 -1 
 
 29 
 
 
 
 
 
 G 
 
 12 
 
 4 
 
 1 
 
 
 
 23 
 
 
 
 
 
 30 
 
 
 
 
 
 14J22 
 
 13 
 
 4 
 
 
 
 53 
 
 
 
 1 
 
 31 
 
 
 
 
 
 
 
 8 
 
 15 
 
 i 
 
 1 
 
 
 36 
 
 
 
 2 
 
 32 
 
 
 
 
 
 
 6 
 
 6 
 
 4 
 
 
 
 16 
 
 
 
 3 
 
 33 
 
 
 
 
 
 
 1 
 
 3 
 
 5 
 
 1 
 
 
 8 
 
 
 
 4 
 
 34 
 
 
 
 
 
 
 
 
 1 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 2 127 
 
 51 
 
 a : 20 
 
 
 
 
 146i 
 
 
 
 
 
 
 1 
 
 
 1 
 
 
 
 
 Fxo. 52. — Correlation in buds of O. rubri- 
 nei~iis,^o. 04113. Thickness of hypan- 
 thium subject, thickness of bud-cone 
 relative. Coefficient of correlation, 
 0.7170 ±0.0259. 
 
 Fig. 53. — Correlation in the buds of 
 O. rtibrin-ertis. No. 04137. Thickness 
 of hypanthium subject, thickness of 
 bud-cone relative. Coefficient of cor- 
 relation, .5397 ± 0.0354. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Xo 
 
 -9 
 
 -8 
 
 -7 
 
 -6 
 
 -5 
 
 -4 
 
 -3 
 
 -2 
 
 -1 
 
 
 
 1 
 
 2 3 
 
 4 
 
 5 
 
 6 
 
 7 
 
 8 
 
 9 ilO 
 
 11 12 
 
 13 
 
 14 
 
 15 
 
 16 
 
 17 
 
 
 
 
 
 
 
 
 16 
 
 17 
 
 IS 
 
 Il9'20l21 
 
 22T23 
 
 24 
 
 25 
 
 26 
 
 27'28'29 30'3l'32'33 3435 
 
 36 37 
 
 38: 39140 '41 
 
 42 
 
 
 
 
 X, 
 
 
 
 
 
 
 
 
 
 
 
 . 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -3 
 
 80 
 
 90 
 
 
 
 2 
 
 1 
 
 
 1 
 
 1 
 
 
 
 
 2 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 7 
 
 
 -2 
 
 90 
 
 95 
 
 
 
 
 2 
 
 
 
 1 
 
 1 
 
 1 
 
 
 1 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 7 
 
 
 -1 
 
 95 
 
 too 
 
 
 1 
 
 
 1 
 
 3 
 
 5 
 
 
 2 
 
 1 
 
 2 
 
 4 
 
 
 
 1 
 
 2 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 22 
 
 
 
 
 100 
 
 105 
 
 
 
 
 1 
 
 
 
 4 
 
 2 
 
 
 2 1 
 
 3 
 
 
 1 ' 1 
 
 
 
 3 
 
 
 
 
 
 
 
 
 
 
 
 18 
 
 
 1 
 
 lOojIlO 
 
 
 
 
 
 1 
 
 1 
 
 1 
 
 1 
 
 2 
 
 2 
 
 1 
 
 1 
 
 1 
 
 
 
 1 
 
 2 
 
 1 
 
 ' 1 
 
 
 
 
 
 
 
 
 1 
 
 
 19 
 
 
 2 
 
 110115 
 
 
 
 
 
 
 
 
 
 
 2 
 
 1 
 
 2 
 
 I 
 
 
 2 
 
 1 2 1 
 
 
 
 
 
 
 
 
 
 
 12 
 
 
 3 
 
 115120 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 4 
 
 120 
 
 |125 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 1 
 
 2 
 
 
 
 
 
 
 
 
 4 
 
 
 5 
 
 125 
 
 130 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 2 
 
 5l4 
 
 7 
 
 7 
 
 6 
 
 4 
 
 8 
 
 HI G 
 
 213174 3 1 3 2 1'2 
 
 1 1 1 1 1 i 1 
 
 
 91 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 1 
 
 
 1 1 
 
 1 
 
 j 
 
 
 
 
 
 Fig. 54. — Correlation in the buds of Oenothera gigas. Length of ovary subject, length 
 of hypanthium relative. Coefficient of correlation, 0.6237 i 3.0432. 
 
 
 
 
 
 1 
 
 Mill 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 Xo 
 
 -9-8-7i-g|-5I-4'-3'-2 
 
 -1 
 
 1 
 
 2 
 
 3 
 
 4 
 
 5 G 
 
 7 8 
 
 9 
 
 10 111 
 
 
 
 
 
 
 
 
 30J31I32 33 34 35 38 37 1 38 
 
 39 
 
 40i41 
 
 42 
 
 43 
 
 44 45 i 46 i 47 
 
 48149150 
 
 
 
 
 
 Xi 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 i 
 
 
 
 
 1 
 
 
 
 
 -3 
 
 85 
 
 90 
 
 
 
 
 
 2 
 
 1 
 
 2 
 
 
 
 
 
 
 
 
 1 
 
 
 1 
 
 
 
 
 
 
 
 7 
 
 
 -2 90 95 
 
 
 1 
 
 
 
 1 
 
 1 
 
 3 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 7 
 
 
 -1 9o|l00 
 
 
 
 
 1 
 
 1 
 
 4 
 
 2 
 
 2 
 
 3 
 
 2 
 
 1 
 
 1 
 
 
 1 
 
 1 
 
 2 
 
 
 
 1 
 
 
 
 
 
 22 
 
 
 100105 
 
 
 
 
 1 
 
 
 
 2 
 
 1 
 
 1 
 
 3 
 
 1 
 
 1 ! 1 1 2 
 
 2 
 
 
 1 
 
 1 
 
 
 
 1 
 
 
 
 18 
 
 
 1 1105110 
 
 
 
 
 1 
 
 
 
 1 
 
 2 
 
 1 
 
 2 
 
 1 
 
 1 
 
 2 
 
 1 
 
 2 
 
 
 2 
 
 
 1 
 
 
 1 
 
 1 
 
 
 19 
 
 
 2 
 
 110H5 
 
 
 
 
 
 
 
 2 
 
 1 
 
 
 2 
 
 1 
 
 1 
 
 
 
 1 
 
 1 
 
 1 
 
 1 
 
 
 
 1 
 
 
 
 12 
 
 
 3 
 
 115120 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 1 
 
 
 4 120125 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 1 
 
 
 1 
 
 1 
 
 
 
 4 
 
 
 5 125130 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 3 
 
 4 
 
 6 
 
 12 
 
 7 
 
 5 
 
 S 
 
 4 
 
 4 
 
 3 
 
 4 
 
 7 
 
 3 
 
 6 
 
 3 
 
 2 
 
 2 
 
 4 
 
 2 
 
 
 91 
 
 — 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 _ 
 
 Fig. 55. — Correlation in the buds of O. gigas. Length of ovary subject, length 
 of bud-cone relative. Coefficient of correlation, 0.5338 ± 0.0506. 
 
50 MUTATIONS, VARIATIONS, AND RELATIONSHIPS O^ THE OENOTHERAS. 
 
 
 
 
 — 
 
 
 
 
 
 
 — 
 
 -~~ 
 
 
 ^~~ 
 
 
 
 
 
 
 
 
 
 
 
 
 — 
 
 
 
 
 X, 
 
 -ti 
 
 -5 
 
 -4 
 
 -3 
 
 -2 
 
 -1 
 
 
 
 1 
 
 2 
 
 3 
 
 4 
 
 5 
 
 6 
 
 7 
 
 8 
 
 9 
 
 10 
 
 11 
 
 
 
 
 
 
 
 38 
 
 39 
 
 40 
 
 41 
 
 42 
 
 43 
 
 44 
 
 45 
 
 46 
 
 47 
 
 48 
 
 49 
 
 50 
 
 51 
 
 521 
 
 53 
 
 54 
 
 o-i 
 
 
 
 
 X, 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 _ 
 
 -3 
 
 85 
 
 90 
 
 
 
 1 
 
 1 
 
 2 
 
 3 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 7 
 
 
 -2 
 -1 
 
 90 
 
 95 
 
 
 1 
 
 1 
 
 3 
 
 
 
 1 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 t 
 
 
 9.5 100 
 
 
 
 
 3 
 
 4 
 
 6 
 
 3 
 
 1 
 
 3 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 22 
 
 
 
 
 loo'ios 
 
 
 1 
 
 
 
 o 
 
 2 
 
 2 
 
 O 
 
 O 
 
 1 
 
 
 I 
 
 1 
 
 1 
 
 1 
 
 1 
 
 
 
 
 
 18 
 
 
 1 
 
 lOSjllO 
 
 
 
 1 
 
 
 
 
 3 
 
 2 
 
 4 
 
 2 
 
 
 1 
 
 1 
 
 2 
 
 1 
 
 1 
 
 
 
 
 
 19 
 
 
 2 
 S 
 
 110115 
 
 
 
 
 
 1 
 
 
 2 
 
 1 
 
 1 
 
 1 
 
 
 o 
 
 1 
 
 o 
 
 
 
 
 
 
 
 12 
 
 
 115120 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 4 
 
 120125 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 1 
 
 
 1 
 
 
 4 
 
 
 5 
 
 125^30 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ~" 
 
 
 
 ~" 
 
 o 
 
 3 
 
 7 
 
 9 
 
 11 
 
 11 
 
 / 
 
 10 
 
 6 
 
 6 
 
 4 
 
 3 
 
 6 
 
 o 
 
 o 
 
 1 
 
 
 
 1 
 
 
 91 
 
 
 
 
 
 
 
 
 
 J_ 
 
 
 
 
 
 
 
 _ 
 
 
 
 __ 
 
 -^ 
 
 _— 
 
 
 
 
 
 
 
 Fig. 56. — Correlation in the buds of Oenothera gigas. Length of 
 ovary subject, thickness of ovary relative. Coefficient of corre- 
 lation, 0.6360 ±0.0421. 
 
 1 
 
 ' 
 
 " 
 
 ~ 
 
 ~ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Xo 
 
 -9 
 
 -8 
 
 -7 
 
 -6| 
 
 -5 
 
 --" 
 
 -3 
 
 _o 
 
 -1 
 
 
 
 1 
 
 2 
 
 3 
 
 4 
 
 5 
 
 6 
 
 7 
 
 8 
 
 9 
 
 10 
 
 11 
 
 
 
 
 
 
 
 
 30 
 
 31 
 
 32 33 1 
 
 34 
 
 35136 
 
 37 
 
 38 
 
 39 
 
 40 
 
 41 
 
 42 
 
 43 
 
 44 
 
 45 
 
 46 
 
 47 
 
 48 
 
 49 
 
 50 
 
 
 
 
 
 X, 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -9 
 
 16 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 -8 
 
 17 
 
 
 
 
 
 O 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 2 
 
 
 -7 
 
 13 
 
 
 
 
 
 
 o 
 
 1 
 1 
 
 
 
 
 
 
 1 
 
 1 
 
 
 
 
 
 
 
 
 
 
 5 
 
 
 
 -6 
 
 19 
 
 
 
 
 
 
 
 1 
 
 
 1 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 4 
 
 
 
 -5 
 
 20 
 
 
 
 
 
 
 1 
 
 2 
 
 1 
 
 2 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 7 
 
 
 
 -4 
 
 21 
 
 
 
 
 
 1 
 
 
 1 
 
 1 
 
 1 
 
 
 1 
 
 
 
 
 1 
 
 
 1 
 
 
 
 
 
 
 
 7 
 
 
 
 -3 
 
 22 
 
 
 1 
 
 
 o 
 
 
 
 1 
 
 
 1 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 6 
 
 
 
 — 2 
 
 23 
 
 
 
 
 
 
 1 
 
 
 2 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 4 
 
 
 
 -1 
 
 24 
 
 
 
 
 
 
 1 
 
 1 
 
 
 
 1 
 
 1 
 
 I 
 
 
 1 
 
 1 
 
 
 
 
 
 
 
 1 
 
 
 8 
 
 
 
 
 
 25 
 
 
 
 
 
 1 
 
 
 1 
 
 1 
 
 
 1 
 
 1 
 
 
 
 
 •■> 
 
 
 1 
 
 
 1 
 
 
 1 
 
 1 
 
 
 11 
 
 
 
 1 
 
 2G 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 1 
 
 
 I 1 1 
 
 
 
 
 1 
 
 
 1 
 
 
 
 6 
 
 
 
 2 
 
 27 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 o 
 
 
 
 3 
 
 28 
 
 
 
 
 
 
 1 
 
 
 1 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 z\ 
 
 
 
 4 
 
 29 
 
 
 
 
 
 
 
 4 
 
 
 
 1 
 
 
 
 1 
 
 
 
 1 
 
 
 
 
 
 
 
 
 7 
 
 
 
 5 
 
 30 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 4 
 
 
 
 C 
 
 31 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 o 
 
 1 
 
 
 
 
 
 
 3 
 
 
 
 7 
 
 32 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 8 
 
 33 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 1 
 
 
 
 1 
 
 
 
 O 
 
 
 
 9 
 
 34 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 2 
 
 
 
 10 
 
 35 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 1 
 
 
 
 11 
 
 36 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 1 
 
 
 
 
 2 
 
 
 
 12 
 
 37 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 13 
 
 38 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 14 
 
 39 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 1 
 
 
 
 15 
 
 40 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 16 
 
 41 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 u 
 
 
 
 17 
 
 42 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 [ 
 
 1 
 
 
 
 3 
 
 4 
 
 6 
 
 12 
 
 7 
 
 5 
 
 9 
 
 4 
 
 4 
 
 3 
 
 4 
 
 7 
 
 3 
 
 6 
 
 3 
 
 2 
 
 2 
 
 4 
 
 2 
 
 
 91 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Fig. 57. — Correlation in the buds of Oenothera gigas. Length of hypanthiura 
 subject, length of bud-cone relative. Coefficient of correlation, 0.5273 
 ±0.0510. 
 
MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE) OENOTHERAS. 5 1 
 
 "^ 
 
 C 
 
 3 
 p 
 
 3 S =» 
 
 ill 
 
 ^ — 5' 
 
 3 o — 
 
 *■ =^ =• 
 
 00 9^ 
 
 •*- 3 ? 
 
 LL ID C 
 
 rr u5 Q< 
 
 0^2, 
 
 ^ =* ^ 
 
 •O 5 
 
 3 S. 
 
 c' 2 
 5 ^ 
 
 re 
 
 3 
 
 ° ? 
 
 3 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 - 
 
 ^ 
 
 Cn 
 
 X 
 
 W 
 
 t« 
 
 ^ 
 
 o 
 
 cc 
 
 GC 
 
 -^ 
 
 c-. 
 
 cn 
 
 M. 
 
 CO 
 
 to «-* 
 
 O 
 
 i .^ 
 
 \ 
 
 1 
 
 1 
 
 1 
 
 C3 
 
 1 
 ^1 
 
 1 
 
 ■i >< 
 
 
 
 
 
 
 
 a. 
 
 .;>. 
 
 S 
 
 Of 
 
 CO 
 
 C*3 
 
 -J 
 
 §? 
 
 CO 
 
 en 
 
 &5 
 
 
 n 
 
 M 
 
 g 
 
 IZ 
 
 'i 
 
 li g 
 
 tc 
 
 i: 
 
 10 
 
 io 
 
 15 
 
 to 
 
 li 
 
 ^ 
 
 00 
 
 ^i 
 
 g 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 X 
 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 - 
 
 
 to 
 
 t 
 
 QO 
 
 
 
 c 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Ki 1 1 
 
 
 
 c 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ^li 
 
 
 cc 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ►- 
 
 
 
 
 
 
 ^ 
 
 " 
 
 
 
 
 M 1 1 
 
 
 ^ 
 
 
 
 
 
 
 
 
 
 
 H- 
 
 
 
 
 
 to 
 
 
 
 " 
 
 
 
 
 >- 
 
 
 
 
 - 
 
 ^ 
 
 
 
 
 1 
 
 
 00 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 - 
 
 
 
 
 - 
 
 
 >- 
 
 ^ 
 
 
 ^ 
 
 - 
 
 - 
 
 t-» 
 
 
 
 
 CO 
 CO 
 
 1 
 
 
 - 
 
 
 
 
 
 
 
 
 
 
 H- 
 
 
 
 
 
 - 
 
 
 
 
 
 - 
 
 t>3 
 
 
 
 - 
 
 ^ 
 
 
 - 
 
 
 
 CO 
 
 .'. 
 
 
 W 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 - 
 
 - 
 
 t— 
 
 
 ts 
 
 " 
 
 
 to 
 
 •u 
 
 ^ 
 
 
 
 
 
 
 £? 
 
 1 
 
 
 C-. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 - 
 
 - 
 
 
 
 
 - 
 
 
 - 
 
 
 - 
 
 1- 
 
 
 
 
 
 I-! 
 
 c 
 
 
 
 00 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 - 
 
 
 
 
 
 to 
 
 - 
 
 
 to 
 
 
 
 
 to 
 
 
 
 
 CO 
 
 - 
 
 
 
 -J 
 
 
 
 
 
 
 
 
 - 
 
 
 
 
 
 
 
 i- 
 
 
 
 C3 
 
 to 
 
 
 
 
 
 
 
 
 
 
 
 CO 
 00 
 
 to 
 
 
 
 ^ 
 
 
 - 
 
 
 
 
 
 
 
 
 
 "^ 
 
 >- 
 
 
 - 
 
 ►- 
 
 
 
 ^ 
 
 " 
 
 
 
 
 
 
 
 
 
 
 
 CO 
 
 Oi 
 
 
 
 <I 
 
 
 
 
 
 - 
 
 
 
 
 
 
 
 
 
 - 
 
 
 
 - 
 
 
 to 
 
 - 
 
 
 
 ^ 
 
 
 
 
 
 
 
 
 ^ 
 
 
 
 c: 
 
 
 
 
 
 
 
 
 
 
 
 »-i 
 
 
 M 
 
 
 
 
 
 
 - 
 
 
 - 
 
 
 - 
 
 
 
 
 
 
 
 
 en 
 
 
 
 .1- 
 
 
 
 
 
 
 
 
 t^ 
 
 t-t 
 
 
 
 
 
 
 
 
 
 
 
 to 
 
 
 
 
 
 
 
 
 
 
 
 05 
 
 
 
 c 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 CO 
 
 ^ 
 
 
 
 ts 
 
 
 
 
 
 
 
 
 
 
 
 - 
 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 oc 
 
 
 
 c 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 cr 
 
 
 ;- 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 o 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 H 
 
 
 ^ 
 
 o 
 
 o 
 
 H- 
 
 o 
 
 c 
 
 »a 
 
 1-^ 
 
 M 
 
 CO 
 
 - 
 
 w£ 
 
 -. ^ 
 
 M 
 
 to 
 
 05 
 
 — 
 
 00 
 
 ^ 
 
 C5 
 
 ^ 
 
 ^ 
 
 .e. c< 
 
 to 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 n 
 
 rr 
 
 
 o 
 
 c 
 
 
 o 
 
 
 vO 
 
 3 
 
 
 1 
 
 o_ 
 
 o 
 
 3 
 
 n 
 
 3 
 
 ft 
 
 
 
 r/) 
 
 ■-1 
 
 O 
 
 C 
 
 
 ..*» 
 
 rr 
 
 p 
 
 r> 
 
 n 
 
 »-»• 
 
 c 
 
 n, 
 
 D 
 
 "I 
 
 r^ 
 
 3 
 
 o 
 
 r*- 
 
 — . 
 
 p 
 
 ^ 
 
 M 
 
 r.- 
 
 o 
 
 C* 
 
 
 
 K* 
 
 c 
 
 a 
 
 3 
 
 0. 
 
 
 O 
 
 Ul 
 
 o 
 
 t/5 
 73 
 
 
 
 4^ 
 
 r. 
 
 
 (^ 
 
 
 ■vl 
 
 Cv 
 
 , 
 
 
 
 ^ 
 
 CI 
 
 
 
 
 H- 
 
 9- 
 
 'S 
 
 o 
 
 o 
 
 
 
 W 
 
 o 
 
 3 
 
 
 o 
 
 re 
 
 re 
 3 
 
 
 a 
 
 Kl 
 
 
 
 — 
 
 
 < 
 
 
 
 re 
 
 
 
 
 
 
 
 
 
 
 ! 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 s 
 
 o 
 
 « 
 
 00 
 
 ^ 
 
 o 
 
 CT 
 
 .£. 
 
 CO to M 
 
 o 
 
 ill 1 
 
 .-. 1 to 1 CO 
 
 1 1 
 
 1 1 
 C3 -4 
 
 1 
 
 CO 
 
 1 
 
 >< 
 
 
 
 
 
 
 
 g 
 
 
 00 
 
 
 
 
 
 CO 
 
 .£k ^ ^ , c^ 
 
 CO 
 00 
 
 £5ig 
 
 CO CO 
 
 CO 
 CO 
 
 CO 
 
 to 
 
 CO 
 
 CO 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 X 
 
 IS 
 
 
 - 
 
 1 
 
 
 
 
 
 
 
 
 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 00 
 
 1 
 
 
 CO 
 
 
 
 
 
 
 
 
 
 
 - 
 
 - 
 
 
 
 
 
 
 
 I— 
 
 
 
 
 
 
 g 
 
 
 
 
 
 1 
 CO 
 
 
 ^ 
 
 
 
 
 
 
 
 
 »Ht 
 
 
 
 
 
 
 H- 
 
 
 
 H- 
 
 ►— 
 
 !-• 
 
 - 
 
 
 - 
 
 
 c 
 
 
 
 « 
 
 1 
 
 to 
 
 1 — 
 
 1 
 
 - 
 
 
 K- 
 
 
 " 
 
 
 
 
 
 
 H* 
 
 
 - 
 
 to 
 
 >!. 
 
 " 
 
 to 
 
 I— 
 
 
 
 
 g 
 
 1 
 
 1 
 
 g 
 
 to 
 
 - 
 
 
 - 
 
 
 
 
 CO 
 
 - 
 
 - 
 
 
 
 
 CO 
 
 ^ 
 
 Cn 
 
 CO 
 
 1-^ 
 
 " 
 
 
 
 
 
 
 
 
 
 
 to 
 
 
 
 
 ^- 
 
 
 
 f- 
 
 
 ►-• 
 
 
 
 - 
 
 - : CT 
 
 
 ^ 
 
 ►— 
 
 
 
 
 
 
 
 c^ ! 
 
 »-• 
 
 — 
 
 
 
 
 
 
 
 
 |-» 
 
 t>9 
 
 1— 
 
 
 - 
 
 - 
 
 lO 
 
 - 
 
 
 •-* 
 
 
 
 
 
 
 
 il5 
 
 to 
 
 
 
 
 - 
 
 
 
 - 
 
 bo 
 
 ^ 
 
 - 
 
 
 - 
 
 - 
 
 - 
 
 1- 
 
 
 
 
 
 
 
 
 
 
 '3 8 
 
 CO 
 
 1 M 
 
 
 
 
 
 
 - 
 
 i-» 
 
 
 
 t— 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 gig 
 
 •^ 
 
 - 
 
 
 
 
 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 u § 
 
 Oi 
 
 - 
 
 
 
 
 
 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 IS 
 
 a 
 
 
 - 
 
 
 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 01 
 
 
 
 M 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 i 
 
 
 to 
 
 a^ 
 
 bO 
 
 to 
 
 CO 
 
 o 
 
 CO 
 
 -^ 
 
 .u. 
 
 CO 
 
 .ii. .£.. 
 
 «o 
 
 en 
 
 *4 
 
 >-* 
 
 to 
 
 0:1 
 
 .c 
 
 CO 
 
 
 
 H- 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
52 MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Xo 
 
 -s 
 
 ~7 
 
 -6 
 
 -5 
 
 -i 
 
 -3 
 
 -2 
 
 -1 
 
 
 
 1 
 
 2 
 
 3 
 
 4 
 
 5 
 
 G 
 
 7 
 
 S 
 
 9 10 
 
 
 
 
 
 
 
 
 2S 
 
 29 
 
 30 
 
 31 
 
 32 
 
 33 
 
 34 
 
 35 
 
 3R 
 
 37 
 
 38 
 
 39 
 
 40 
 
 41 
 
 42 
 
 43 
 
 44 
 
 45 
 
 46 
 
 
 
 
 
 X, 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 - G 
 
 38 
 
 
 
 
 
 
 1 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 2 
 
 
 
 -5 
 
 39 
 
 
 
 
 
 1 
 
 1 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 3 
 
 
 
 -4 
 
 40 
 
 
 1 
 
 
 
 1 
 
 
 3 
 
 1 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 7 
 
 
 
 ~I> 
 
 41 
 
 
 
 
 
 
 3 
 
 
 o 
 
 3 
 
 
 
 
 
 
 
 
 
 
 
 
 
 9 
 
 
 
 _o 
 
 42 
 
 
 
 
 
 
 
 2 
 
 o 
 
 1 
 
 2 
 
 2 
 
 
 
 1 
 
 
 
 
 
 
 
 
 ^1 1 
 
 
 -1 
 
 43 
 
 — 
 
 
 
 
 1 
 
 
 1 
 
 2 
 
 o 
 
 2 
 
 
 1 
 
 1 
 
 
 
 
 
 
 
 
 
 11 
 
 
 
 144 
 
 
 
 
 
 o 
 
 
 1 
 
 2 
 
 1 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 7 
 
 
 
 1 
 
 45 
 
 
 
 
 
 
 
 1 
 
 
 2 
 
 
 2 
 
 3 
 
 1 
 
 
 1 
 
 
 
 
 
 
 
 10 
 
 
 
 n 
 
 46 
 
 
 
 
 
 
 
 
 
 1 
 
 
 2 
 
 o 
 
 1 
 
 
 
 
 
 
 
 
 
 6 
 
 
 
 o 
 
 47 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 U 
 
 1 
 
 1 
 
 
 
 
 
 
 
 6 
 
 
 
 4 
 
 48 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 1 
 
 
 
 1 
 
 1 
 
 
 
 
 
 
 4 
 
 
 
 5 
 
 49 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 1 
 
 1 
 
 
 
 
 
 
 3 
 
 
 
 G 
 
 30 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 4 
 
 
 
 
 1 
 
 
 
 
 G 
 
 
 
 7 
 
 51 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 2 
 
 
 
 
 
 
 
 O 
 
 
 
 s 
 
 52 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 1 
 
 
 O 
 
 
 
 y 
 
 53 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 1 
 
 
 
 10 
 
 54 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 11 
 
 55 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ^ 
 
 
 
 1 
 
 
 
 
 
 
 I 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 ■J 
 
 7 
 
 S 11 
 
 13 
 
 
 
 S 
 
 7 
 
 7 
 
 7 
 
 G !4 
 
 
 
 o 
 
 
 
 1 
 
 
 91 
 
 
 
 
 ! 
 
 
 
 
 
 
 
 _ 
 
 
 1 
 
 
 
 
 
 
 
 
 
 Fig. 60. — Correlation in the buds of Oenothera gigas. Thickness of ovary 
 subject, thickness of hvpanthium relative. Coefficient of cor- 
 relation, 0.8385 ±0.0210. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Xo 
 
 -4 
 
 -3 
 
 -2-1 
 
 1 
 
 2 
 
 3 4 
 
 5 6 7 
 
 
 
 
 
 
 
 X,U4 
 
 -3 
 
 -o 
 
 -1 
 
 
 
 1 
 
 o 
 
 3 
 
 4 
 
 5 
 
 G 
 
 7 
 
 
 
 
 
 
 
 
 
 85 
 
 90 
 
 95 lOO'lOS 
 
 no 
 
 115 
 
 120 125[l30 13.5 140 
 
 
 
 
 
 
 
 
 S.3 
 
 90 
 
 95 
 
 100 
 
 105 
 
 llOHa 
 
 120 
 
 125 
 
 130 135 
 
 110 
 
 
 
 
 
 
 
 
 00 
 
 95 
 
 looiosno'us 
 
 120|l2.5130!l35U0H5 
 
 
 
 
 
 
 
 
 90 
 
 9.5 
 
 100105 
 
 no 
 
 115 
 
 120 
 
 125 
 
 130 
 
 135 
 
 140 
 
 145 
 
 
 
 
 
 Xi 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Xi 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -8i28 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 -6 
 
 38 
 
 
 
 1 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 2 
 
 
 
 -7i29 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -5 
 
 39 
 
 
 
 
 1 
 
 2 
 
 
 
 
 
 
 
 
 
 
 3 
 
 
 
 -G|30 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -4 
 
 40 
 
 
 
 
 3 
 
 2 
 
 2 
 
 
 
 
 
 
 
 
 
 t 
 
 
 
 -5)31 
 
 
 
 
 1 
 
 1 
 
 
 
 
 
 
 
 
 
 
 3 
 
 
 
 -3 
 
 41 
 
 
 
 
 1 
 
 2 
 
 5 
 
 1 
 
 
 
 
 
 
 
 
 9 
 
 
 
 -4 
 
 32 
 
 
 
 1 
 
 2 
 
 1 
 
 
 1 
 
 1 
 
 
 
 
 
 
 
 7 
 
 
 
 —2 
 
 42 
 
 
 
 
 
 2 
 
 4 
 
 4 
 
 1 
 
 
 
 
 
 
 
 11 
 
 
 
 -3 
 
 33 
 
 
 
 
 2 
 
 
 
 1 
 
 2 
 
 
 
 
 
 
 
 8 
 
 
 
 -1 
 
 43 
 
 
 
 
 
 1 
 
 G 
 
 1 
 
 o 
 
 
 
 
 
 
 
 11 
 
 
 
 -2 
 
 34 
 
 
 
 
 
 4 
 
 
 2 
 
 
 1 
 
 
 
 
 
 
 11 
 
 
 
 
 44 
 
 
 
 
 
 3 
 
 1 
 
 1 
 
 1 
 
 1 
 
 
 
 
 
 
 i 
 
 
 
 -1 
 
 35 
 
 
 
 
 
 3 
 
 
 O 
 
 3 
 
 
 
 
 
 
 
 13 
 
 
 
 1 
 
 45 
 
 
 
 1 
 
 
 
 1 
 
 3 
 
 2 
 
 b 
 
 1 
 
 
 
 
 10 
 
 
 
 
 
 3G 
 
 
 1 
 
 
 
 2 
 
 
 1 
 
 
 
 
 
 
 
 
 6 
 
 
 
 o 
 
 46 
 
 
 
 
 1 
 
 1 
 
 
 1 
 
 1 
 
 1 
 
 
 1 
 
 
 
 
 G 
 
 
 1 
 
 37 
 
 
 
 2 
 
 1 
 
 1 
 
 
 2 
 
 
 1 
 
 
 
 
 
 
 8 
 
 
 
 3 
 
 47 
 
 
 1 
 
 
 
 
 1 
 
 
 1 
 
 2 
 
 1 
 
 
 
 
 
 G 
 
 
 
 2 
 
 38 
 
 
 
 
 
 2 
 
 
 1 
 
 1 
 
 1 
 
 
 1 
 
 
 
 
 7 
 
 
 
 4 
 
 48 
 
 
 
 1 
 
 
 1 
 
 1 
 
 
 
 1 
 
 
 
 
 
 
 4 
 
 
 
 3 
 
 39 
 
 
 
 
 
 
 
 
 2 
 
 2 
 
 1 
 
 
 1 
 
 
 
 7 
 
 
 
 5 
 
 49 
 
 
 
 
 
 
 2 
 
 
 
 1 
 
 
 
 
 
 
 o 
 
 
 
 4 
 
 40 
 
 
 
 
 
 
 
 1 
 
 1 
 
 
 
 
 
 1 
 
 
 7 
 
 
 
 G 
 
 50 
 
 
 
 
 
 
 2 
 
 
 1 
 
 
 1 
 
 
 1 
 
 1 
 
 
 G 
 
 
 
 5 
 
 41 
 
 
 
 
 
 1 
 
 
 
 
 4 
 
 
 
 
 
 
 6 
 
 
 
 7 
 
 51 
 
 
 
 
 
 1 
 
 1 
 
 
 
 
 
 
 
 
 
 2 
 
 
 
 6 
 
 42 
 
 
 
 
 
 
 2 
 
 1 
 
 
 1 
 
 
 
 
 
 
 4 
 
 
 
 8 
 
 52 
 
 
 
 
 
 
 
 
 1 
 
 
 1 
 
 
 
 
 
 2 
 1 
 
 
 
 7 
 
 43 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 9 
 
 53 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 8 
 
 44 
 
 
 
 
 
 
 
 
 1 
 
 
 1 
 
 
 
 
 
 2 
 
 
 
 10 
 
 54 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 9 
 
 45 
 
 
 
 
 
 
 
 
 
 
 , 
 
 
 
 
 
 
 
 
 
 11 
 
 55 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 1 
 
 
 
 10 
 
 4G 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 I 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 3 
 
 i 
 
 15 
 
 20 
 
 12 
 
 11 
 
 10 
 
 3 
 
 1 
 
 1 
 
 1 
 
 
 91 
 
 
 
 
 
 
 1 
 
 3 
 
 7 
 
 15 
 
 26 
 
 12 
 
 n 
 
 10 
 
 3 
 
 1 
 
 1 
 
 1 
 
 
 91 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Fig. 61. — Correlation in the buds of Oenothera gigas. 
 Thicknessof ovary subject, thickness of bud-cone 
 relative. Coefficient of correlation, 0.4750 ± 
 0.0747. 
 
 Fig. 62. — Correlation in the buds of Oenothera gigas. 
 Thickness of hypanthium subject, thickness of 
 bud-cone relative- Coefficient of correlation, 
 0.4822 ±0.0543. 
 
MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 5;^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Xn 
 
 L-iiUi 
 
 -10 
 
 -9 
 
 -8 
 
 1 
 
 -6 
 
 -5 
 
 -4 
 
 -3 
 
 -2 
 
 -1 
 
 
 
 1 
 
 
 
 3 
 
 4 
 
 5 
 
 6 
 
 i 
 
 8 
 
 9 
 
 10 
 
 11 
 
 12 
 
 13 
 
 
 
 
 
 
 
 |1-1 
 
 15 
 
 2C|17 
 
 18 
 
 19 
 
 20 
 
 21 
 
 00 
 
 23 
 
 24 
 
 25 
 
 2r.l27 
 
 28 ' 29 
 
 30 
 
 31 
 
 32 
 
 3:j 
 
 34 
 
 35 
 
 36 
 
 37 
 
 38 
 
 39 
 
 
 
 
 Xi 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 I 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -3 
 
 65 
 
 70 
 
 
 
 
 
 
 
 
 
 
 1 
 
 1 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 3 
 
 
 _0 
 
 70 
 
 75 
 
 
 1 
 
 
 
 3 
 
 1 
 
 
 
 ] 
 
 
 
 ] 
 
 1 
 
 ] 
 
 1 
 
 1 
 
 
 
 1 
 
 
 
 
 1 
 
 
 
 
 
 
 
 15 
 
 
 -1 
 
 ■TC 
 
 SO 
 
 
 
 1 
 
 
 
 1 
 
 
 
 
 1 
 
 
 
 3 
 
 2 
 
 2 
 
 
 
 4 
 
 
 1 
 
 
 
 4 
 
 ] 
 
 2 
 
 1 
 
 
 1 
 
 
 
 
 30 
 
 
 
 
 SO 
 
 85 
 
 
 
 
 
 
 1 
 
 
 
 
 
 2 
 
 
 
 3 
 
 
 
 3 
 
 1 
 
 
 
 
 
 
 6 
 
 3 
 
 2 
 
 4 
 
 
 
 
 
 
 33 
 
 
 1 
 
 So ao 
 
 
 
 1 
 
 1 
 
 
 1 
 
 
 
 
 1 
 
 4 
 
 
 1 
 
 3 
 
 1 
 
 
 1 
 
 
 
 4 
 
 ] 
 
 
 
 
 1 
 
 1 
 
 
 
 
 
 25 
 
 
 
 
 90 95 
 
 
 
 
 
 
 
 
 2 
 
 1 
 
 
 1 
 
 1 
 
 2 
 
 3 
 
 3 
 
 
 1 
 
 
 
 
 1 
 
 2 
 
 2 
 
 1 
 
 
 1 
 
 1 
 
 
 22 
 
 
 3 
 
 95 
 
 100 
 
 
 
 
 
 
 
 
 
 
 ] 
 
 
 
 3 
 
 
 3 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 8 
 
 
 4 
 
 100 
 
 105 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 2 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 3 
 
 
 5 
 
 105 
 
 110 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 I 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 1 
 
 3 i 4 1 4 
 
 5 
 
 4 
 
 10 
 
 3 
 
 8 
 
 14 
 
 9 
 
 12 
 
 6 
 
 4 2 
 
 10 
 
 11 
 
 5 
 
 9 
 
 8 
 
 2 
 
 1 
 
 1 
 
 1 
 
 
 140 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Fig. 63. — Correlation in the buds of Oenothera lata. Length of ovary subject, length of 
 hypanthium relative. Coefficient of correlation, 0.0944 ± 0.0565. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Xo 
 
 -11 
 
 -10 
 
 -9 
 
 -8 
 
 ~7 
 
 -6 
 
 -5 
 
 -4 
 
 -3 
 
 —2 
 
 -1 
 
 
 
 1 
 
 2 
 
 3 
 
 4 
 
 5 
 
 6 
 
 7 
 
 8 
 
 9 
 
 10 
 
 
 
 
 
 
 
 
 19 
 
 20 
 
 21 
 
 22 
 
 23124 
 
 25 
 
 26 
 
 27 
 
 28 
 
 29 
 
 30 
 
 31 
 
 32 
 
 33 
 
 34 
 
 35 
 
 36 
 
 37 
 
 38 
 
 39 
 
 40 
 
 
 
 
 X, 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -3 
 
 65 
 
 70 
 
 
 
 
 
 
 2 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 3 
 
 
 —0 
 
 70 
 
 75 
 
 
 1 
 
 
 
 
 
 
 
 
 
 4 
 
 2 
 
 1 
 
 
 
 2 
 
 1 
 
 
 
 
 
 
 
 
 
 15 
 
 
 -1 
 
 75 
 
 80 
 
 
 
 
 
 
 1 
 
 
 
 n 
 
 
 
 3 
 
 3 
 
 5 
 
 
 
 4 
 
 4 
 
 
 1 
 
 
 
 
 
 
 
 30 
 
 
 
 
 80 
 
 85 
 
 
 
 
 
 1 
 
 1 
 
 
 
 
 2 
 
 5 
 
 G 
 
 3 
 
 
 
 
 
 4 
 
 
 1 
 
 1 
 
 
 
 1 
 
 
 
 O-J 
 
 
 1 
 
 85 
 
 90 
 
 
 
 
 
 
 
 
 
 
 1 
 
 4 
 
 1 
 
 1 
 
 4 
 
 5 
 
 
 
 1 
 
 .■; 
 
 
 2 
 
 
 
 1 
 
 
 25 
 
 
 
 
 90 
 
 95 
 
 
 
 
 
 
 
 
 
 2 
 
 2 
 
 
 
 
 
 
 5 
 
 3 
 
 4 
 
 1 
 
 1 
 
 
 
 2 
 
 
 
 •^■t 
 
 
 3 
 
 95 100 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 2 
 
 2 
 
 3 
 
 1 
 
 
 
 S 
 
 
 4 
 
 100105 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 1 
 
 
 1 
 
 
 
 
 
 3 
 
 
 5 
 
 105110 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 1 
 
 4 
 
 6 
 
 4 
 
 11 
 
 14 
 
 14 
 
 11 
 
 9 
 
 10 
 
 11 
 
 lii 
 
 5 
 
 7 
 
 4 
 
 5 
 
 4 
 
 4 
 
 1 
 
 ^140 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 i 
 
 
 Fig. 64. — Correlation in the buds of Oenothera lata. Length of ovary subject, length 
 of bud-cone relative. CoefiScient of correlation, 0.6382 rb 0.0338. 
 
 
 
 
 
 
 
 ' 
 
 
 — \ 
 
 
 
 
 
 
 
 
 
 
 
 Xo 
 
 -4 
 
 -3 
 
 h2 
 
 -1 
 
 
 
 1 
 
 2 
 
 3 
 
 4 
 
 5 
 
 G 
 
 
 
 
 
 
 
 
 34 
 
 35 
 
 36 
 
 o7 
 
 38 
 
 3'J 10 
 
 41 42 
 
 43 
 
 41 
 
 
 
 
 X, 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -3 
 
 05 
 
 70 
 
 
 1 
 
 
 2 
 
 
 
 
 
 
 
 
 
 
 3 
 
 
 -2 
 
 70 
 
 75 
 
 
 
 2 
 
 5 
 
 3 
 
 2 
 
 2 
 
 1 
 
 
 
 
 
 
 15 
 
 
 -1 
 
 75 
 
 80 
 
 
 
 
 8 
 
 8 
 
 4 
 
 2 
 
 4 
 
 3 
 
 1 
 
 
 
 
 30 
 
 
 
 
 80 
 
 8.5 
 
 
 
 1 
 
 5 
 
 5 
 
 4 
 
 9 
 
 6 
 
 1 
 
 1 
 
 1 
 
 
 
 33 
 
 
 1 
 
 85 
 
 90 
 
 
 1 
 
 1 
 
 1 
 
 4 
 
 7 
 
 5 
 
 3 
 
 2 
 
 1 
 
 
 
 
 25 
 
 
 2 
 
 90 
 
 95 
 
 
 
 2 
 
 3 
 
 2 
 
 4 
 
 4 
 
 3 
 
 2 
 
 2 
 
 
 
 
 22 
 
 
 3 
 
 95 
 
 100 
 
 
 
 
 
 
 1 
 
 2 
 
 2 
 
 2 
 
 
 
 1 
 
 
 8 
 
 
 4 
 
 100 
 
 105 
 
 
 
 
 
 
 
 2 
 
 1 
 
 
 
 
 
 
 3 
 
 
 5 
 
 1C5 
 
 110 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 2 
 
 6 
 
 24 
 
 22 
 
 22 
 
 2fi 
 
 21 
 
 10 
 
 5 
 
 1 
 
 1 
 
 
 ^40 
 
 
 
 
 
 
 
 
 
 
 
 -J 
 
 
 
 
 
 
 
 
 
 Fio. 65. — Correlation in the buds of Oenoiliera 
 lata. Length of ovary subject, thickness 
 of ovary relative. Coefi&cient of correla- 
 tion, 0.3568 ±0.0497. 
 
54 MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 „ 
 
 CI 
 
 - 
 
 CO 
 
 -"T 
 
 ^ 
 
 »c 
 
 T 
 
 o 
 
 CO 
 
 oo 
 
 -V 
 
 o 
 
 C-1 
 
 o 
 
 ^ 
 
 CI 
 
 o 
 
 ;^ 
 
 iC 
 
 c 
 
 00 
 
 01 
 
 '- 
 
 - 
 
 -■ 
 
 
 ^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ■"a- 
 
 CO 
 
 CO 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 " 
 
 
 
 
 
 
 
 
 
 i-H 
 
 
 
 CO 
 
 u 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -^ 
 
 - 
 
 
 - 
 
 
 
 
 
 
 01 
 
 
 
 
 
 
 lO 
 
 
 
 C-1 
 
 CO 
 
 
 
 -- 
 
 
 
 
 rH 
 
 - 
 
 
 01 
 
 - 
 
 -H 
 
 
 -- 
 
 CO 
 
 
 r-. 
 
 
 f-l 
 
 
 
 
 CI 
 
 - 
 
 
 tH 
 
 
 
 l^ 
 
 
 
 - 
 
 B 
 
 
 
 
 - 
 
 
 CO 
 
 - 
 
 CO 
 
 01 
 
 TP 
 
 
 (M 
 
 T 
 
 M 
 
 -r 
 
 CI 
 
 
 
 CO 
 
 ■^ 
 
 
 Ol 
 
 01 
 
 " 
 
 
 
 
 
 o 
 
 
 
 o 
 
 ?5 
 
 
 
 
 
 
 
 - 
 
 
 - 
 
 
 
 CO 
 
 o 
 
 -- 
 
 ci 
 
 -H 
 
 CI 
 
 -- 
 
 CI 
 
 - 
 
 CI 
 
 ira 
 
 ei 
 
 
 -- 
 
 
 
 
 o 
 
 CO 
 
 
 
 T 
 
 oo 
 
 
 
 
 
 M 
 
 - 
 
 
 
 
 - 
 
 -- 
 
 -- 
 
 rr 
 
 CO 
 
 -- 
 
 CO 
 
 
 - 
 
 -^ 
 
 T 
 
 CO 
 
 
 
 
 
 
 f-l 
 
 
 oo 
 
 C-1 
 
 
 
 N 
 
 ?T 
 
 
 
 - 
 
 
 
 
 i-» 
 
 - 
 
 - 
 
 01 
 
 
 
 -< 
 
 -- 
 
 -- 
 
 
 
 
 eo 
 
 -H 
 
 
 -^ 
 
 
 
 
 
 
 
 -T" 
 
 
 
 r 
 
 52 
 CI 
 
 
 '^ 
 
 
 
 
 
 
 
 
 -- 
 
 _. 
 
 ^ 
 
 -^ 
 
 
 
 
 
 
 
 
 
 
 -H 
 
 
 
 
 
 
 
 lO 
 
 
 
 X 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 lO 
 
 o 
 
 1^ 
 
 00 
 
 a 
 
 g 
 
 01 
 
 ?5 
 
 CO 
 
 
 CI 
 
 ^ 
 
 ^ 
 
 s 
 
 C5 
 CI 
 
 o 
 
 CO 
 
 CO 
 
 CI 
 
 CO 
 
 CO 
 CO 
 
 CO 
 
 J3 
 
 CD 
 CO 
 
 eo 
 
 00 
 CO 
 
 s 
 
 
 
 
 
 
 
 X 
 
 C-1 
 
 :=: 
 
 o 
 
 3S 
 1 
 
 00 
 
 1 
 
 1 
 
 1 
 
 1 
 
 1 
 
 CO 
 
 1 
 
 CI 
 
 1 
 
 T 
 
 o 
 
 1— 1 
 
 c-1 
 
 CO 
 
 Tp 
 
 « 
 
 o 
 
 t-- 
 
 00 
 
 CS 
 
 o 
 
 HJ 
 
 01 
 
 eo 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 r-» 
 
 1 
 
 £ 
 
 o 
 o 
 
 •"— 
 
 
 c 
 
 
 
 ■*-< 
 
 t> 
 
 
 O 
 
 _jj- 
 
 
 o 
 
 
 
 
 o 
 
 -H 
 
 
 •^ 
 
 > 
 
 .'7\ 
 
 
 
 
 o 
 
 •d 
 
 3 
 
 ■*-; 
 
 3 
 
 
 ^ 
 
 _, 
 
 Xt 
 
 H 
 
 2 
 
 o 
 
 a 
 
 
 p 
 
 
 ■" 
 
 j:: 
 
 3 
 
 n 
 
 
 
 rt 
 
 -M 
 
 "1 
 
 -*-< 
 
 u 
 
 3 
 
 
 rt 
 
 >. 
 
 r^ 
 
 a. 
 
 
 
 Q 
 
 u 
 
 , 
 
 >. 
 
 
 
 n 
 
 
 
 J3 
 
 u 
 
 1 
 
 w 
 
 
 
 ^ 
 
 1 
 
 hll 
 
 
 ■M 
 
 
 a; 
 
 t-l 
 
 3 
 'o 
 
 « 
 
 
 
 
 
 
 
 
 Ih 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 " 
 
 01 
 
 t-( 
 
 CO 
 
 'S- ^ 
 
 LO 
 
 rr 
 
 S " 
 
 00 
 
 ^ 35 
 
 01 
 
 CS 
 
 TP 
 
 Ol 
 
 O 
 
 r-t 
 
 LO 
 
 Ci CO 
 
 Ol 
 
 -^ 
 
 - 
 
 rH 
 
 
 o 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 o 
 
 o 
 
 
 
 
 
 
 
 
 
 
 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 « 
 
 
 
 05 
 
 CO 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 r-4 
 
 
 f-H 
 
 
 
 -H 
 
 
 
 rH 
 
 
 
 
 
 
 TT 
 
 
 
 CO 
 
 ac 
 
 CO 
 
 
 
 
 
 
 
 
 
 
 
 
 
 CO 
 
 
 
 
 
 
 - 
 
 
 
 
 
 
 
 
 
 
 T 
 
 
 
 t- 
 
 S 
 
 
 
 
 
 
 
 
 
 l-H 
 
 w* 
 
 
 t-4 
 
 
 
 Ol 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 lO 
 
 
 
 o 
 
 S 
 
 
 
 
 
 
 
 
 rH 
 
 
 f-t 
 
 
 
 
 -^ 
 
 
 
 -* 
 
 
 
 
 
 
 
 
 
 
 
 
 T 
 
 
 
 LO 
 
 LO 
 
 CO 
 
 
 
 
 
 
 rH 
 
 
 (-• 
 
 
 
 
 C^ 
 
 " 
 
 
 
 
 
 
 r-l 
 
 
 
 
 'H 
 
 
 
 
 
 
 t^ 
 
 
 
 •^ 
 
 CO 
 
 
 
 
 
 
 
 
 
 
 rH 
 
 
 
 rH 
 
 - 
 
 »-t 
 
 
 
 
 
 
 
 
 
 .^ 
 
 
 
 
 
 LO 
 
 
 
 CO 
 
 CO 
 CO 
 
 
 
 
 l-H 
 
 
 »-» 
 
 -H 
 
 T^ 
 
 ^ 
 
 CI 
 
 
 »-H 
 
 rH 
 
 
 
 -H 
 
 
 
 -H 
 
 - 
 
 
 -H 
 
 
 
 
 
 -^ 
 
 
 •^ 
 
 
 
 N 
 
 CI 
 
 CO 
 
 
 
 
 
 
 
 - 
 
 -■ 
 
 
 ^ 
 
 f-H 
 
 
 -H 
 
 Ol 
 
 -^ 
 
 
 
 
 
 
 
 
 e^ 
 
 
 
 ^ 
 
 
 
 Z, 
 
 
 
 ^H 
 
 CO 
 
 
 
 
 
 -- 
 
 -H 
 
 l-t 
 
 
 
 1-H 
 
 r^ 
 
 
 
 
 
 
 
 
 
 « 
 
 
 F-( 
 
 »-i 
 
 
 rH 
 
 
 
 
 C' 
 
 
 
 O 
 
 ^ 
 
 
 
 - 
 
 
 
 
 - 
 
 
 
 
 
 
 
 
 - 
 
 
 
 
 
 - 
 
 01 
 
 
 Ol 
 
 -H 
 
 
 
 
 
 C2 
 
 
 
 1 
 
 5i 
 
 
 
 
 
 
 
 
 
 
 
 
 - 
 
 
 '- 
 
 Ol 
 
 
 
 -- 
 
 C^ 
 
 
 
 CO 
 
 .H 
 
 
 
 
 
 
 = 
 
 
 
 01 
 
 1 
 
 00 
 
 
 
 -- 
 
 
 
 
 
 
 
 ^ 
 
 
 
 Cl 
 
 '- 
 
 
 Ol 
 
 rH 
 
 
 C-1 
 
 CO 
 
 
 - 
 
 
 
 
 
 
 
 "^ 
 
 
 
 CO 
 
 1 
 
 s 
 
 
 
 
 
 ^ 
 
 
 
 - 
 
 
 
 
 
 01 
 
 -^ 
 
 -- 
 
 01 
 
 - 
 
 
 C-? 
 
 r-t 
 
 " 
 
 -- 
 
 
 
 
 
 
 
 -r 
 
 
 
 •V 
 
 ^ 
 
 
 " 
 
 
 
 -^ 
 
 
 
 
 - 
 
 
 
 Ol 
 
 
 
 01 
 
 
 
 - 
 
 
 
 Ol 
 
 - 
 
 
 
 
 
 
 
 - 
 
 
 
 in 
 1 
 
 s 
 
 
 
 
 
 
 
 
 
 
 -- 
 
 
 
 
 ^ 
 
 - 
 
 - 
 
 
 
 
 
 
 
 
 
 
 
 
 
 rr 
 
 
 
 CD 
 1 
 
 C4 
 
 
 
 
 
 
 -H 
 
 
 
 - 
 
 
 
 
 01 
 
 
 
 
 
 
 rH 
 
 
 
 rH 
 
 
 
 
 
 
 
 cr 
 
 
 
 1 
 
 CO 
 CI 
 
 
 
 
 
 
 
 
 
 
 -- 
 
 - 
 
 
 
 '^ 
 
 
 
 
 
 
 " 
 
 
 
 
 
 
 
 
 
 TT 
 
 
 
 00 
 
 1 
 
 a 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -- 
 
 
 
 
 
 
 
 
 
 - 
 
 
 
 ? 
 
 s 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 o 
 
 
 
 o 
 
 s 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 O 
 
 
 
 
 c; 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -H 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ^ 
 
 
 
 X 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 i« 
 
 >o 
 
 CD 
 
 t* 
 
 00 
 
 Ci 
 
 ?. 
 
 01 
 
 CI 
 
 S 
 
 Ol 
 
 s 
 
 ci 
 
 ^1 
 
 ?3 
 
 Oi 
 01 
 
 u 
 
 CO 
 
 Ol 
 
 CO 
 
 IS 
 
 CO 
 
 S 
 
 s 
 
 S 
 
 00 
 CO 
 
 C5 
 CO 
 
 
 
 
 
 
 
 X 
 
 Ol 
 
 1-4 
 
 O 
 
 Ci 
 
 GO 
 1 
 
 1 
 
 CO 
 1 
 
 »fO 
 
 1 
 
 T 
 
 CO 
 
 1 
 
 CI 
 
 1 
 
 1 
 
 o 
 
 ^H 
 
 M 
 
 CO 
 
 T 
 
 in 
 
 CD 
 
 h- 
 
 CO 
 
 Ol 
 
 2 
 
 rH 
 
 rH 
 
 CO 
 
 r^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 P 
 
 -n 
 
 5 
 
 vO 
 
 /^ 
 
 -»■ 
 
 -*-' 
 
 
 
 O 
 
 
 
 c 
 
 o 
 
 
 '2 O 
 
 ^ '-J 
 
 c i3 
 
 3 u 
 
 _D 3 
 
 O 
 
 - "3 
 
 3 O 
 
 " 
 
 
 
 %-> 
 
 
 
 
 >-i\ 
 
 
 a> 
 
 a 
 
 
 
 V 
 
 
 Vm 
 
 
 
 
 
 , 
 
 
 u 
 
 -4-1 
 
 n 
 
 
 s; 
 
 
 vr 
 
 s: 
 
 o 
 
 ^r 
 
 3 
 
 
 
 w 
 
 
MUTATIONS, VARIATIONS, AND REI.ATIONSHIPS OF THE OENOTHERAS. 55 
 
 
 
 
 Xc 
 
 -2 
 
 -1 
 
 
 
 1 
 
 2 
 
 3 
 
 4 
 
 5 
 
 6 
 
 7 
 
 8 
 
 
 
 
 
 
 
 
 80 
 
 85 
 
 90 
 
 95 
 
 lOfl 
 
 105 
 
 lie 
 
 115 
 
 12C 
 
 12) 
 
 130 
 
 
 
 
 
 
 
 
 85 
 
 90 
 
 95 
 
 IOC 
 
 105 
 
 IXC 
 
 115 
 
 120 
 
 125 
 
 130 
 
 135 
 
 
 
 
 
 X, 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -11 
 
 19 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 
 -10 
 
 20 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -9 
 
 •21 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -8 
 
 22 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -7 
 
 23 
 
 
 O 
 
 O 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -6 
 
 24 
 
 
 o 
 
 3 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -5 
 
 25 
 
 
 1 
 
 3 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -4 
 
 26 
 
 
 
 C 
 
 4 
 
 1 
 
 
 
 
 
 
 
 
 
 11 
 
 
 
 ^3 
 
 27 
 
 
 
 3 
 
 8 
 
 3 
 
 
 
 
 
 
 
 
 
 14 
 
 
 
 — •> 
 
 28 
 
 
 
 4 
 
 6 
 
 O 
 
 o 
 
 
 
 
 
 
 
 
 14 
 
 
 
 -1 
 
 29 
 
 
 
 1 
 
 7 
 
 3 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 30 
 
 
 
 
 O 
 
 4 
 
 
 o 
 
 
 
 
 
 
 
 
 
 
 1 
 
 31 
 
 
 
 
 3 
 
 3 
 
 
 
 
 
 
 
 
 
 
 
 
 2 
 
 32 
 
 
 
 1 
 
 1 
 
 4 
 
 
 1 
 
 1 
 
 
 
 
 
 
 
 
 
 3 
 
 33 
 
 
 
 
 
 8 
 
 
 o 
 
 
 
 
 
 
 
 
 
 
 4 
 
 34 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 5 
 
 35 
 
 
 
 
 1 
 
 1 
 
 
 o 
 
 
 
 
 
 
 
 
 
 
 G 
 
 36 
 
 
 
 
 
 
 
 o 
 
 1 
 
 
 
 
 
 
 
 
 
 7 
 
 37 
 
 
 
 
 
 1 
 
 
 3 
 
 
 
 
 
 
 
 
 
 
 8 
 
 38 
 
 
 
 
 
 
 
 o 
 
 2 
 
 
 
 
 
 
 
 
 
 9 
 
 39 
 
 
 
 
 
 
 1 
 
 
 1 
 
 
 1 
 
 
 1 
 
 
 
 
 
 10 
 
 40 
 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 6 
 
 24 
 
 33 30] 
 
 24 
 
 15 
 
 6 
 
 
 
 1 
 
 
 
 1 
 
 
 40 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Fig. 68. — Correlation in the buds of Oenottiera 
 lata. Length of bud-cone subject, thickness 
 of bud-cone relative. CoefiBcient of correlation, 
 0.8366 ±0.0171. 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 ^ 
 
 
 
 
 
 Xo 
 
 -3 
 
 -2 
 
 -1 
 
 
 
 1 
 
 
 
 3 
 
 4 
 
 
 
 
 
 
 
 
 26 
 
 27 
 
 28 
 
 29 
 
 30 
 
 31 
 
 32 
 
 33 
 
 
 
 
 
 X, 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -4 
 
 I34 
 
 
 1 
 
 1 
 
 
 
 
 
 
 
 
 2 
 
 
 
 -3 
 
 35 
 
 
 1 
 
 
 
 2 
 
 1 
 
 
 
 
 
 
 6 
 
 
 
 —2 
 
 36 
 
 
 1 
 
 7 
 
 9 
 
 4 
 
 2 
 
 1 
 
 
 
 
 24 
 
 
 
 -1 
 
 37 
 
 
 
 3 
 
 8 
 
 6 
 
 5 
 
 
 
 
 
 22 
 
 
 
 
 
 38 
 
 
 
 
 4 
 
 6 
 
 10 
 
 2 
 
 
 
 
 22 
 
 
 
 1 
 
 39 
 
 
 2 
 
 
 3 
 
 5 
 
 13 
 
 2 
 
 1 
 
 
 
 26 
 
 
 
 2 
 
 40 
 
 
 
 1 
 
 2 
 
 5 
 
 5 
 
 7 
 
 1 
 
 
 
 21 
 
 
 
 3 
 
 41 
 
 
 
 
 
 1 
 
 4 
 
 4 
 
 1 
 
 
 
 10 
 
 
 
 4 
 
 42 
 
 
 
 
 
 2 
 
 1 
 
 
 2 
 
 
 
 5 
 
 
 
 5 
 
 43 
 
 
 
 
 
 
 
 
 
 1 
 
 
 1 
 
 
 
 6 
 
 44 
 
 
 
 
 
 
 
 1 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 5 
 
 14 
 
 28 
 
 30 40 1 
 
 17 
 
 5 
 
 1 
 
 
 140 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Fig. 69. — Correlation in the buds of Oen- 
 othera lata. Thickness of ovary sub- 
 ject, thickness of hypanthium rela- 
 tive. Coefficient of correlation, . 6400 
 ±0.0336 
 
 
 
 
 XJ-2 
 
 -1 
 
 
 
 1 
 
 
 
 3 
 
 4 
 
 6 
 
 i 
 
 8 
 
 
 
 
 
 
 
 
 80 
 
 85 
 
 90 
 
 95 
 
 100 
 
 105 
 
 no 
 
 115 
 
 120 
 
 125 
 
 130 
 
 
 
 
 
 
 
 
 85 
 
 90 95 
 
 100105 
 
 110 
 
 115 
 
 120 
 
 125 
 
 130 
 
 135 
 
 
 
 
 
 Xi 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -4 
 
 34 
 
 
 1 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -3 
 
 35 
 
 
 
 
 
 3 
 
 1 
 
 
 
 
 
 
 
 
 
 6 
 
 
 
 —2 
 
 36 
 
 
 4 
 
 8 
 
 7 
 
 4 
 
 1 
 
 
 
 
 
 
 
 
 24 
 
 
 
 -1 
 
 37 
 
 
 
 5 
 
 8 
 
 3 
 
 6 
 
 
 
 
 
 
 
 
 22 
 
 
 
 |.38 
 
 
 
 4 
 
 5 
 
 4 
 
 8 
 
 1 
 
 
 
 
 
 
 
 22 
 
 
 
 1 !39 
 
 
 
 3 
 
 ^ 
 
 10 
 
 1 
 
 8 
 
 
 
 
 
 
 
 26 
 
 
 
 2 40 
 
 
 1 
 
 
 
 3 
 
 6 
 
 4 
 
 2 
 
 3 
 
 
 
 
 
 
 21 
 
 
 
 3 
 
 41 
 
 
 
 
 2 
 
 
 2 
 
 3 
 
 3 
 
 
 
 
 
 
 10 
 
 
 
 4 
 
 42 
 
 
 
 
 
 
 
 2 
 
 
 
 
 1 
 
 
 
 
 5 
 
 
 
 a 
 
 43 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 1 
 
 
 
 6 
 
 44 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 6 24 
 
 33 
 
 30 
 
 24 
 
 IS 
 
 6 
 
 
 
 1 
 
 
 
 1 
 
 
 140 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 Xj-2-1 
 
 
 
 1 
 
 2 
 
 3 1 4 1 5 
 
 6 
 
 7 
 
 S 
 
 
 '~~ 
 
 
 
 
 
 
 80|85 
 
 90 
 
 95 
 
 100 
 
 105 
 
 110|115 
 
 120 
 
 125 
 
 130 
 
 
 
 
 
 
 
 
 85 
 
 90 
 
 95 
 
 100105 
 
 110 
 
 115120 
 
 125 
 
 131 
 
 135 
 
 
 
 
 
 f^ 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 -3 
 
 26 
 
 
 1 
 
 3 
 
 
 
 
 1 
 
 
 
 
 
 
 
 5 
 
 
 
 p2i27 
 
 
 3 
 
 4 
 
 3 
 
 3 
 
 1 
 
 
 
 
 
 
 
 
 14 
 
 
 
 -i|28 
 
 
 2 
 
 8 
 
 10 
 
 4 
 
 1 
 
 3 
 
 
 
 
 
 
 
 28 
 
 
 
 1 29 
 
 
 
 5 
 
 8 
 
 8 
 
 c 
 
 1 
 
 *> 
 
 
 
 
 
 
 30 
 
 
 j 1 
 
 30 
 
 
 
 4 
 
 6 
 
 14 
 
 10 
 
 4 
 
 
 
 
 
 
 
 40 
 
 
 1 2 
 
 31 
 
 
 
 
 6 
 
 1 
 
 4 
 
 'i 
 
 1 
 
 
 
 
 
 
 17 
 
 
 ?, 32 
 
 
 
 
 
 
 2 
 
 1 
 
 1 
 
 
 1 
 
 
 
 
 5 
 
 
 
 4 
 
 33 
 
 
 
 
 
 
 
 
 
 
 
 
 1 
 
 
 1 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 6 
 
 24 
 
 33 
 
 30 
 
 24 
 
 1.5 
 
 6 
 
 
 
 1 
 
 
 
 1 
 
 140 
 
 
 1 
 
 
 
 1 
 
 
 
 
 
 
 
 
 
 1 
 
 
 Fig. 70. — Correlation in the buds of Oenothera 
 lata. Thickness of ovary subject, thickness 
 of bud-cone relative. Coefficient of corre- 
 lation, 0.5839 ±0.0376. 
 
 Fig. 71. — Correlation in the buds of Oenothera 
 lata. Thickness of hypanthium subject, 
 thickness of bud-cone relative. Coefficient 
 of Icorrelation, 0.5549 ± 0.0394. 
 
56 MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 
 
 PEDIGREE-CULTURES OF OENOTHERA GRANDIFLORA, O. BIENNIS, AND 
 
 O. CRUCIATA. 
 
 In order to test the mutative capacity of other evening-primroses, seeds 
 were obtained from correspondents in various parts of the world and as many 
 as 40 sowings were made in one year from 0. biennis. In only one strain, 
 however, did anything come to light which would be of interest in connection 
 wnth the chief purpose of these investigations. 
 
 Parental individuals were selected and verified by Dr. N. L. Britton in 1903, 
 and from the seeds furnished by them the plants were grown which furnished 
 material for the descriptive diagnosis published in a previous paper (Mac- 
 Dougal, Vail, ShuU & Small, 1905). This is not the species growing wild in 
 Europe and cited by De Vries in his "Mutationstheorie. " 
 
 Inflorescences of 4 separate individuals were inclosed in transparent paper 
 bags in such manner as to secure pure fertilization and the seeds preserved in 
 separate packages. Sowings were made in pans of sterilized soil, in accord- 
 ance with the usual custom, early in January, 1905. 
 
 A number of seedlings were seen in which the first foliage-leaf and all suc- 
 ceeding leaves were very much narrower than the parental type. Sixteen 
 individuals of this kind were found and preserved, appearing in the progeny 
 of all of the 4 parents represented. An exact count of the number was kept 
 only in the case of one parent. No. 8.17, in which 4 of these aberrants were 
 included in a progeny of 669 individuals and formed very nearly 0.6 per cent 
 of the entire number. 
 
 The aberrants retained their property of forming narrow leaves, which w'ere 
 slightly paler in color than the parental type, and appeared to taper gradually 
 at the base into long petioles. One aberrant of 8.17 gave off a lateral branch 
 at the base of the stem which developed broader leaves resembling those of 
 the parent. With the completion of development, the form of the capsules 
 and the entire habit of this branch were in exact duplication of O. biennis. 
 
 On July 2 one of the aberrants opened a flower which also showed some 
 departures from the normal, being characterized by the extremely long sta- 
 mens and by the elongated capsule, which was of equal diameter throughout 
 and did not taper to a point or toward the tip as in its parental form. The 
 first 4 flowers which were brought to maturity on a lateral branch were inclosed 
 in a bag and self -pollinated. The seeds formed were allowed to mature and 
 these were sown as soon as practicable in order to determine the constancy 
 of the new form. 
 
 The adult plant has a main stem varying from 80 to 150 cm. in height, and 
 branches freely from the base upward. The lower branches are long, slender 
 and assurgent, while the upper ones are much shorter. Numerous secondary 
 
MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 57 
 
 branches arise during the latter part of the season. The stems are nearly 
 terete below, but are slightly channeled above. 
 
 The stem-leaves are 10 to 16 cm. long, and not more than i cm. in width 
 at the middle, obscurely pubescent on both surfaces, thickish, yellowish green, 
 obscurely and irregularly denticulate, with a heavy, broad midvein. The 
 leaves are linear, and taper to the long acuminate tip and to the margined 
 petiole. 
 
 The bracts are narrowly linear and 3 cm. long. The conic portion of the 
 bud is about 12 mm. long and half that in diameter, pale yellow, and barely 
 tapering in the apical portion, being finely pubescent. The free erect tips are 
 irregularly acuminate and 2 to 4 mm. long. 
 
 The hypanthium is 25 to 35 mm. long and exceeds the cal^TC-lobes by half. 
 
 The filaments are 7 to 9 mm. long, stout, and the anthers 4 mm., slender. 
 The pistil is slightly shorter or longer than the stamens, measuring 4 to 6 
 mm., and this species is therefore self-pollinated like the parental form. The 
 stigmatic lobes are 4 to 5 mm. long and spreading. 
 
 The capsules are 25 to 35 mm. long, 4 to 5 mm. in diameter, generally 
 curv^ed, sparingly pubescent, bright green and shining, scarcely tapering to the 
 apex, and, in general, much thinner than the parental type. The ovary is 
 about half or less than half the length of the mature capsule. 
 
 The sepals are 3 mm. long, the petals firm, 15 to 16 mm. long and 15 mm. 
 wide, being crenately emarginate (1905 c, p. 17). 
 
 A small package of the lot of seeds from which the mutant was obtained 
 originally were sent to INIiss Elizabeth Billings, who sowed them in her green- 
 house at Woodstock, Vermont, in March, 1905, and later transferred them to 
 her experimental garden. The 499 plants thus obtained were kept under 
 observation by Miss Billings and Professor F. S. Lee, of Columbia University, 
 with the result that aberrant individuals were found which were identical 
 with those obtained during the previous September in the New York Botanical 
 Garden. 
 
 The development of the newly arisen form has been followed from Septem- 
 ber, 1904, until the present time. The first individual discovered came into 
 bloom on July 2, 1905, having been cultivated under glass during the preced- 
 ing Mdnter. Seeds were matured 5 or 6 wrecks later, and sowings have been 
 made from them, with the result that the aberrant type was found to produce 
 a progeny which contained not more than 12 per cent of the atypic form of 
 individuals, the remainder being of the parental type and unchanged. Hybrid- 
 izations between the aberrant form and the parental form gave the same 
 result, while the aberrant type when crossed with lamarckiana and with ruh- 
 rinervis gave a progeny resembling in general constituency those produced 
 by typical biennis. 
 
58 MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 
 
 A similar aberrant of 0. cniciata has been found in many of the cultures of 
 that species. Much attention has been devoted to these forms, both as to 
 habits of growth and anatomy. The only parallel occurrences are those 
 which are to be found in other genera in which atypical individuals affected by 
 the enzymatic diseases may give a progeny in which the normal form of indi- 
 viduals predominates. The vigorous growth of the atypic individuals and 
 the comparative regularity with which it appears are well worthy of remark. 
 If these forms were found growing in the open, they might easily be taken as 
 belonging to a species apart from 0. biennis unless their progeny was tested 
 in pure cultures. 
 
 The pedigreed cultures of 0. grandiflora made in 1906 from purely fertilized 
 seeds of the previous year included about 1 500 plants. More than one season 
 of such cultures is necessary to secure definite results that may be depended 
 upon. The observations have been carried so far at the New York Botan- 
 ical Garden and the Desert Laboratory, however, as to warrant the assertion 
 that this species presents a complex progeny analogous to that of lamarckiana, 
 in which two well-defined mutants are readily recognizable by reason of their 
 striking diflferences from the parental form. 
 
 ORIGIN OF FIXED FORMS BY HYBRIDIZATION. 
 
 That the cross-fertilization of two forms may result in the production of a 
 unitypic or polytypic progeny in which the parental qualities appear in a 
 mosaic is well known. Components of a native flora have been suspected to 
 be of hybrid origin in a few instances, and by succeeding experimental tests 
 have been synthetized from the parents. The practice prevailing among 
 taxonomists of ascribing a hybrid origin to a newly discovered form, which, 
 in outward anatomical characters, is between two know^n species, is extremely 
 pernicious and is not justified by facts obtained in cultural work. The best 
 grounds for such a conclusion are to be found when two species alone occur 
 in a region, and the appearance of a third is attributed to hybridization ; but 
 even here the supposition that a mutation may have ensued is allowable. 
 
 The hybrid 0. lamarckiana X 0. cruciata (form with long hypanthium and 
 slender bud), which was described in the previous publication by the authors, 
 proves to be a fixed form, and as it sets seeds freely and is self-fertilizing, it is 
 in every respect an independent species. Several hundreds of seedlings were 
 grown from purely fertilized capsules ripened in 1904, with the result that all 
 w^ere seen to conform to the parental hybrid type in every particular. The 
 two parental forms do not meet in their native habitats, and hence this form 
 could not have arisen in a state of nature. 
 
 The reciprocal of this cross, that is, 0. cruciata X 0. lamarckiana, consists of 
 three forms, one of which is indistinguishable from the 0. biennis cruciata 
 which has been received from various European gardens. 
 
MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 59 
 
 BUD-SPORTS— VEGETATIVE SALTATIONS. 
 
 One of the most interesting and least understood phases of heredity is that 
 in which the tract of embryonic tissue constituting a bud may develop and 
 put into external realization a set of characters wholly different from those of 
 the remainder of the individual. A large number of such instances have been 
 seen by gardeners, horticulturists, and farmers, although exact observations 
 on the hereditary qualities of such mutant branches are almost wholly want- 
 ing. In some cases the departure from the type does not affect the entire 
 branch, and it may not be made manifest until an advanced stage of its devel- 
 opment, causing some of the leaves or perhaps some of the flowers to show 
 atypic characters. When such partial vegetative saltation ensues it gen- 
 erally results in converting a lateral section of the branch or inflorescence into 
 the new form, and has been termed "sectorial variation" byDeVries. To 
 designate it beyond danger of misapprehension, however, it should be known 
 as sectorial bud-mutation. 
 
 The cultures carried on in the experimental garden during 1905 were char- 
 acterized by two remarkable bud-sports. 
 
 The hybrid 0. lamarckiana X {0. lamarckiana X 0. cruciata) comprised 
 nearly a score of recognizable forms in the seedling stage. About two of 
 each type were transplanted to the experimental garden in May, 1905. Two 
 of the types included were apparently different from each other only in the 
 character that one of them bore flowers with broad petals resembling those of 
 lamarckiana, while the other was furnished with cruciate flowers, but with the 
 petals broader than the typical cruciata and quite as long as those of lamarck- 
 iana. It was evident that the two forms constituted an illustration of 
 Mendelian combinations, alike in all particulars except as to the characters 
 imited in the form and size of the petals. 
 
 One of the individuals with the broadly cruciate flowers bore a branch near 
 the base of the stem, on which were formed only flowers of the broadly-petaled 
 lamarckiana type. Some of these were purely fertilized and seeds preserved 
 for testing. These were duly sown in December, 1905, under glass, and 
 representative individuals brought to bloom in the open air in July, 1906. 
 
 When they came into bloom in July, 1906, it was found that the branch- 
 sport came entirely true to the type which it represented. With this it is 
 interesting to note that the purely fertihzed seeds of the main stem bearing 
 cruciate flowers gave a progeny which contained some individuals bearing 
 flowers with broad petals, while the other strain, which bore only broad 
 petals, came entirely true to its type without deviation. No exact numerical 
 count was made, but it seemed evident that the cruciate petal is dominant 
 over the cordate, and that the plants described above exemplify the develop- 
 ment of a sport on a plant of the first generation in which the recessive char- 
 acter appeared and when extracted came true to the recessive character. 
 
6o MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 
 
 De Vries (1900, p. 86) described a case of vegetative mutation in a hybrid 
 between Veronica longifolia, which has a blue flower, and V . alba with a 
 white flower. The hybrid has a blue flower, but several cases were noted in 
 which the entire buds or portions of inflorescences produced white flowers. 
 These white sports, when self-fertilized, produced white flowers only, in the 
 preliminary tests cited. 
 
 Similar vegetative splitting is known in Datura hybrids, being seen by 
 Naudin in 3 individuals only, while sports of this character of Brassica, Raph- 
 aniis, Abies, Anagallis, Hclianthemum, Zea, and Cytisiis are well known. 
 Nearly all other sports of hybrids are clearly of an atavistic nature, being 
 more or less direct reversions, sometimes rehearsing the juvenile characters 
 of one of the parents of the hybrids. 
 
 A second instance was offered by a form known as 0. ammophila. This 
 material was grown from seeds obtained from Professor W. O. Focke, which 
 had been collected on the coast near Bremen, Germany. 
 
 A large number of seeds were sown in sterilized soil in the propagating 
 house on February 10, 1905. A month later the seedlings obtained were 
 examined and a dozen representing the widest visible range of variation were 
 transferred to pots, while the remainder of the culture was discarded. With 
 successive repottings the number of individuals was reduced to 7, which were 
 transplanted to the experimental garden in the latter part of May, at which 
 time no differences of moment were noticed. During the summer 3 of the 
 specimens sent up shoots, while the other 4 formed dense rosettes with a few 
 short lateral branches. 
 
 About the middle of August the more advanced individuals came into 
 bloom. No careful examination of them had been made up to this time, 
 but it was now seen that one of them had formed a lateral branch at the 
 base of the main stem, which, by the spread of its branches, the shape and 
 expansion of its leaves, and its general vigorous growth had become the larger 
 member of the shoot. This branch, in the form and behavior of all of its 
 organs, including flowers and fruits, was an exact reproduction of the O. 
 biennis which has formed the basis of cultures under that name in the New 
 York Botanical Garden. Branches of the sport and of typical biennis were 
 submitted to several botanists, with the result that they were found to be 
 indistinguishable by anatomical characters. 
 
 The main stem, which had branched in the usual manner, had been crowded 
 from its natural upright position and had assumed a half- recumbent position, 
 having the appearance of a lateral branch. 
 
 The suggestion lies near at hand, that O. ammophila is a hybrid derivative 
 of O. biennis, and that the vegetative mutation is simply one of reversion, after 
 the manner of examples cited above. No positive evidence upon the origin of 
 ammophila is at hand, however. 
 
o. 
 
 PLATE 9. 
 
 Ot-nothera aminophila with bud-sport. The narrow -leaved decumbent stock represents the typical parent, 
 while the more vigorous upright branch with broader leaves is the mutant branch which represents 
 Oenothera biennis. 
 
MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 6 1 
 
 Purely fertilized seeds of the mutant branch and of the main stem were 
 obtained, and these were sown late in September, 1905. Those of the main 
 stem came true to the type of the main stem, that is, to 0. ammophila. All 
 of the progeny obtained from the purely guarded seeds of the sport conformed 
 strictly to the biennis type (plate 9). 
 
 THE INDUCTION OF MUTATIONS. 
 
 One of the most important problems confronting the investigation in this 
 subject is that of the localization of the mutations in the Hfe-history of the 
 plant. If we could fix upon the exact stage in which the direct changes 
 occurred, we then would be in a position to examine the protoplasts con- 
 cerned with a view to ascertaining what changes ensue in the chromosomes in 
 connection with saltations in inheritance. Furthermore, opportunity would 
 also be offered for attempts to determine the factors which operate as stim- 
 uli, or loosing agents in setting the mutations free. 
 
 A theoretical consideration of the subject seemed to indicate that the 
 changes constituting the mutations which give rise to atypic seedlings took 
 place in a stage previous to the reduction divisions in the embryo-sac or in 
 the pollen mother-cell. It was planned, therefore, to subject these structures 
 to the action of chemical agents at a time before fertilization had occurred. 
 The attempt was made to secure two forms of stimulation by using some solu- 
 tions of high osmotic value, and other mineral compounds which are stimu- 
 lative in low concentrations. A desideratum in such experimentation was to 
 use plants in which a large number of ovules were to be found in one ovarial 
 cavity. The solutions were injected into the ovaries by means of a physician's 
 hypodermic syringe. Operations of this character were carried out with 
 Begonia rotundifolia and with a species of Cleome, but with negative results, 
 and injections were also made into the central placenta of a number of flowers 
 of Abutilon abutilon, but in all cases at a very late stage, and so far no results 
 have been obtained with this plant. It is being used in a new series of experi- 
 ments, however, which, with improved technique, may be expected to furnish 
 some interesting material. 
 
 Attention was next turned to several species of Oenothera which were 
 being cultivated in pure pedigreed strains in the experimental garden. 
 Solutions of copper sulphate of i to 400,000 were injected into ovaries of 
 lamarckiana immediately previous to pollination and the pistils were purely 
 pollinated. The capsules formed from ovaries treated in this way were vari- 
 ously distorted as a result of the wounding, but a large crop of perfect seeds 
 was matured, which were sowed in September, 1905. A careful census of the 
 seedlings late in November made it appear that in the small number of mutants 
 present the usual mutants of this species occurred in normal proportions. 
 
62 MUTATIONS, VARIATIONS, AN'D RELATIONSHIPS OF THE OENOTHERAS. 
 
 A number of radium pencils were taken from a series used by Dr. Gager in 
 his investigations, and were fastened in an inflorescence in such manner that a 
 deleterious influence must have been exerted on some ovules, while in other 
 cases the effect must have been stimulative if the reactions of seedlings of 
 other plants may be taken to offer a fair analogy. The crop of seeds matured 
 in the treated ovaries was sown in September, 1905, and late in November a 
 census showed that only 20 normal specimens of laniarckiana were present 
 and that 2 mutants were also included. The rate of mortality was thus seen 
 to be twice as great in the type as in the mutants in this single test. 
 
 Similar tests were made with 0. biennis. Solutions of magnesium sulphate 
 in distilled water were used without any noticeable departure in the composi- 
 tion or behavior of the progeny arising from the seeds. Poisonous solutions 
 containing i part of zinc sulphate in 500 of distilled water were also used. 
 The seeds produced in the ovaries which had received this treatment contained 
 atypic forms of the kind seen in ordinary cultures in apparently the cus- 
 tomary proportion of about one in 200. In addition however, a single rosette 
 was found which differed widely from any known type, and of this form, which 
 was recognizably different from the parental type in many qualities, some of 
 the differences were plainly apparent even in the earliest leaves of the seedlings. 
 These differences have become accentuated in the adult plant. 
 
 The parental form has been under observation for five years in cultures 
 and in a wild condition. An aberrant form, which appears to be eversporting, 
 has been previously figured, and while this form appeared in the injected 
 or treated seeds in a normal proportion, yet the newest aberrant has not been 
 seen elsewhere. The probability must be taken into account that it may be 
 a mutant of rare occurrence, the cycle of which came within the experiments ; 
 but in either case it is plainly a mutant, and it only remains to be seen whether 
 or not it was induced by the action of the zinc solution. The presumption 
 seems to favor such a conclusion. Seeds of this form were harvested in 
 August, 1906, and sown immediately, with the result that the characters of 
 the new form were found to be fully transmissible, the first generation of the 
 progeny being duplicates of the parent within the limits of fluctuating varia- 
 bility. The entire plant is characterized by a much deeper green color than 
 the parental form and the leaves are slightly curled and twisted, owing to 
 inequalities of growth, and it reaches maturity quite early in the season. 
 
 In addition to the crop of guarded seeds harvested from some of the 
 branches, the available remainder were collected and sown in the greenhouse of 
 the Dessrt Laboratory, with the result that this second generation of the deriv- 
 ative was found to conform exactly in every individual to the derivative type. 
 Either the mutant does not intercross with the parent, although the branches 
 were in contact, or if a cross has occurred the derivative qualities are dominant. 
 This matter may be determined by the time this paper finds its way through 
 
PLATE 10. 
 
 B 
 
 A. Raimannia and iiuluced derivatives. The three rows of plants in the background include some large 
 
 normal individuals and small mutants grown from seed taken from a capsule which had been treated 
 with calcium nitrate. The remainder are mutants of the second generation. 
 
 B. Individuals grown from seeds taken horn one capsule. The large rosette is of the parental type, the 
 
 smaller are derivatives, two of which have come into bloom at the same age and are scarcely so high 
 as the normal rosette. 
 
MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 63 
 
 the press. Among the large progeny grown from guarded seeds was only one 
 which showed any marked variation from the blunt rounded leaves of the 
 rosette of the derivative, but this variation did not extend to the foliar organs 
 developed later. 
 
 The radium pencils which were later used on 0. lamarckiana were affixed to 
 an inflorescence in such a manner that the radium coatings were 15 to 25 mm. 
 from developing flower-buds. The corollas of many were so retarded that 
 they failed to open and fell off prematurely. At greater distances develop- 
 ment of the ovary proceeded but slowly and normal size was not reached. 
 Perfect seeds were formed in many of them, however, and these when sown 
 gave the normal frequency of the aberrant mentioned above, and which is not 
 to be confused with the one induced by the stimulative action of zinc sulphate. 
 
 Some decisive results were also obtained from Raimannia odorata, a member 
 of a separate genus of the evening-primrose family from Patagonia. During 
 the first season (1905) injections of the ovaries were made with several sub- 
 stances, with the result that an atypic form identical in all cases was found in 
 seeds from ovules that had been treated in various ways. Two such mutants 
 were secured from seeds of an ovary that had been treated with a 10 per cent 
 sugar solution, 10 from one that had been injected with a solution of calcium 
 nitrate i part to 1000 of distilled water, and one was also found in the 
 progeny from seeds taken from a capsule which had been exposed to the 
 action of a radium pencil. 
 
 In all of these injections ordinary jointed metal syringes were used, and the 
 water was from a single distillation in a copper still, so that no special reliance 
 may be placed upon the purity of the solution injected into the ovary. 
 
 Seeds from capsules which had been treated were harvested in June, 1905, 
 and were sown early in July. The atypic derivatives could be recognized as 
 soon as the cotyledons were fully expanded, and no skill was needed for their 
 detection. The parental form bore leaves villous hairy, or with ciliate mar- 
 gins, while the mutant was entirely and absolutely glabrous. The leaves of 
 the parent generally have an unbalanced linear growth of the margins by 
 which the blades become fluted. The excess of growth in the mutant lies 
 along the midrib and the margins become re volute. The leaves of the mutant 
 show much less expansion than those of the parent, the lower ones being 
 narrowly linear, 5 to 6 cm. long and 3 to 4 mm. wide, while the upper ones are 
 lanceolate linear, 3 to 4 cm. long or even longer, the mutant being much the 
 narrower. The parental type is of a marked biennial habit and near the close 
 of the season the internodes formed are extremely short, which results in a 
 dense rosette. The mutant does not make a rosette, by reason of the fact 
 that its stem does not alter its rate of elongation, but proceeds at a uniformly 
 rapid rate, thus presenting a leaf-stem which appears to be perennial in the 
 climate in which the experiments were performed. So rapidly did develop- 
 
64 MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 
 
 ment proceed that one of the atypic individuals opened a flower-bud on No- 
 vember 27, 1905, and seeds were harvested before the beginning of the year. 
 Other lots of seeds were obtained throughout the year and several sowings 
 were made, with the result that the atypic form was found to transmit its con- 
 stellation of characters to its offspring. Among the fully atypic individuals, 
 of which a few hundred have been grown, one was found with wide leaves like 
 the parental form, another with ciliate leaves, and another with undulate 
 leaves. This activation of parental characters is one fully illustrated by 
 lamarckiana and its derivatives and shows that the treatment has simply 
 thrown certain parental characters into a state of latency and awakened 
 others with which the parental characters are mutually exclusive as to external 
 manifestation. These individuals did not afford a transition or intergrading 
 series between the derivative and the parental form, however (plate 10). 
 
 In the summer of 1906 a second series of injections was made, with the result 
 that capsules treated with zinc sulphate, i part to 2000 parts of distilled 
 water, yielded seeds which produced progeny inclusive of the atypic form 
 described above, and also some other combinations which it has not been 
 possible to follow in successive generations. The group, however, bears a 
 general resemblance in relationship to that of the lamarckiana mutants. 
 
 An injection with calcium nitrate, i part to 1000 of distilled water, was 
 without effect and the progeny were all of the parental type. It was remark- 
 able, however, that a treatment with this substance in the previous season 
 had secured some atypic forms, probably due to some opportune condition in 
 the experiment not yet understood. Furthermore, the plants of the progeny 
 of the first treatment which were apparently normal yielded seeds which gave 
 a few atypic forms, indicating that the effect of the first treatment had been 
 more or less permanent. Confirmation of this important matter has not yet 
 been obtained, however. Then, again, an injection of ovaries in 1906 with the 
 distilled water subject to impurity from the still and from the syringe, as men- 
 tioned above, also resulted in a progen}^ in which some atypic forms* were 
 found to occur. So far, then, as experience with this plant is afforded it is to 
 be seen that a variety of agents act in inducing discontinuous variation in the 
 progeny and that in one instance the variability was carried to the third 
 generation, as far as that part of the test has been extended. The atypic 
 forms transmit their qualities perfectly from generation to generation, and 
 the third generation now in hand are like the first from which they came 
 originally. 
 
PLATE 11 . 
 
 :-^ '-^ 
 
 ?s... 
 
 B 
 
 A. Ounotlicra i^randirtoni, seedling alioul two months after yenninatioii. 
 K. (Jenotlicra grandiriora, ro.sette five months ohl. 
 
IDENTITY OF EVENING-PRIMROSES. 
 
 Prepared by Anna Murray Vail. 
 
 In view of the extent to which the evening-primroses are being used to 
 obtain experimental evidence upon questions of heredity at the present time, 
 it has been deemed important to make a systematic study of the group. To 
 this end seeds have been obtained from correspondents in various parts of 
 America and from European botanical gardens in which these plants appear 
 to be extensively cultivated. 
 
 In the few years in which the cultures have been under way it has become 
 plainly apparent that a much larger number of species are native to America 
 than has been supposed hitherto. The series of pedigree-cultures made from 
 them, using carefully guarded seeds, shows that botanists have customarily 
 grouped many elementary species under the name of O. biennis. Furthermore, 
 as has been noted on page 8, none of these appear to coincide with the form 
 cultivated in Europe under this name. 
 
 Several of these elementary species occur on Long Island, a number of them 
 collected by Mr. E. P. Bicknell deserving mention, and requiring special study 
 to determine their relationship to the common 0. biennis. 
 
 Another group of closely related but apparently distinct elementary species 
 came from the west shore of Lake Champlain. There, in waste ground at the 
 roadsides about Plattsburg, fairly typical 0. biennis is to be found. A small- 
 flowered and very characteristic plant abounds on inland sand-dunes, in 
 sandy clearings along the Ausable River, and on the railroad embankments 
 near Bluff Point. In the Ausable River woods it grows associated with 0. 
 cruciata, and a supposed hybrid between the two was observed. Occasionally 
 a larger-flowered form on the edge of woods was observed and still another on 
 the gravelly shores of the lake. These types, with the exception of the hybrid 
 of which no mature seed was secured, were grown in the New York Botanical 
 Garden, where they reproduced in every particular the characters that distin- 
 guished them in the field. They will require more extended study for the 
 determination of their specific relationships. 
 
 The more important facts concerning the anatomy and distribution of a 
 few forms have been obtained and descriptions of O. grandiflora, 0. simsiana, 
 O. oakesiana, 0. parviflora, and O. muricata are here presented. It is proposed 
 to reserve discussion of the large number of other distinct forms until their 
 life-history and hereditary qualities shall have been more definitely ascer- 
 tained. 
 
 65 
 
66 MUTATIONS, VARIATIONS, AND REIvATlONSHIPS OF THE OENOTHERAS. 
 
 The use here of the name Oenothera for species belonging to the genus 
 Onagra (Tournefort, Adanson, Spach) is according to the decision of Dr. J. N. 
 Rose in a recent paper (1905), where he points out, and with good reason, that 
 Oenothera biennis should be considered as the type of the genus. The descrip- 
 tions in each case have been made from living plants. 
 
 OENOTHERA GRANDIFLORA Aiton 
 
 Oenothera grandiflora Ait. Hort. Kew., 3: 2, 1789. 
 
 Onosiiris aciiminita Raf. Fl. Ind., 96, 1817 (?). 
 
 Oenotliera grandiflora P Sims. Curtis's Bot. Mag. 46: pi. 2068, 1819. 
 
 Oenothera biennis var. grandiflora Lindl. Bot. Regist. 19: pi. 1604, 1833 (?). 
 
 Onagra grandiflora (Ait.) Vail. Torreya, 5: 9, 1905. 
 
 Onagra vulgaris Spach. Nouv. ann. mus. Paris, 4: 353, 1835. (Reprint 33, 1835.) 
 
 In part. 
 
 Seedling about 2 months old. — Rosette loose, spreading, 7 to 12 cm. in diameter; outer 
 leaves 4 to 7 cm. long, 2 to 3 cm. wide; blades oblong, or oblong-spatulate, nearly glabrous, 
 irregularly blotched with dull red, broader above the middle, rounding or acutish and 
 shallowly undulate toothed at the apex, tapering at the more deeply toothed base into the 
 margined petiole ; the inner blades oblong, acutish, and soon becoming more deeply toothed 
 at the base (plate 11, a) . 
 
 Seedling about 5 vionths old. — Rosette rather loose, spreading flat on the ground, about 
 30 cm. in diameter; outer leaves 12 to 15 cm. long, 3 to 4 cm. wide; blades oblong-spatulate, 
 broadest above the middle, acutish at the apex, tapering at the irregularly pinnatifid- 
 toothed base into a long, broad petiole, obscurely puberulent near the outer margin above 
 and beneath on the white midvein and veins, bright dark blue-green, irregularly blotched 
 throughout with red, which color also appears sometimes on the wide white midvein 
 (plate II, b). 
 
 Adult plant. — Plant 1.5 to 3 m. in height, branching from the base upward, the terminal 
 portion of the main stem bearing a cluster of short branches with a profusion of flowers. 
 Stems stout, often reaching a diameter of 6 cm. at base, nearly terete below, slightly angled 
 and channeled above, or sulcate, the epidermal tissues becoming detached on the lower 
 part of the main stem, which is hard and woody, clothed in the upper portion with short, 
 spreading, somewhat varyinglymuricate pubescence; leaves 10 to 20 cm. long; blades ovate- 
 lanceolate, glabrate, with appressed scattered hairs on the veins, denticulate, being more 
 deeply and irregularly toothed in the basal portion, tapering to a short margined petiole, 
 bright green and shining above, paler beneath; terminal rosette-like cluster of the inflores- 
 cence symmetrical ; flowers very abundant and fragrant ; bracts lanceolate, 3 to 5 cm. long, 
 acuminate, abruptly tapering into a short petiole, or sessile ; conic portion of bud 3 to 4 cm . long, 
 6 to 7 mm. in diameter at base, very slender, tapering from base to apex, sparingly pubes- 
 cent, thin, dotted with small red spots, the erect free tips very slender, setaceous, 8 to 10 mm. 
 long; hypanthium 4.5 to 5.5 cm. long, very slender, much longer than the reflexed calyx- 
 lobes that usually cohere in pairs at the tips; ovary i cm. long, or less, slender, petals firm, 
 3.5 to 4 cm. long, and about 4 cm. wide, more or less emarginate, wedge-shaped at base; 
 filaments 2.5 cm long, very slender; anthers slender, i cm. long; pistil much longer than 
 stamens, and projecting from the flower and from the unopened buds late in the season; 
 stigmatic lobes 10 to 12 mm. long; capsule 3 to 3.5 cm. long, 8 mm. in diameter in thickest 
 portion, green and shining, glabrous or with a few scattered hairs, four-angled and tapering 
 from base (plate 12). 
 
 Alabama. — Earle's Landing and Dixie Landing near Tensaw, Tracy, No. 8001 in the 
 herbarium of the New York Botanical Garden. 
 
PLATE 12. 
 
 Oenothera grandiflora. 1, stem-leaf; 2, leaf from young rosette; 3, bract; 4, bud; 5, flower 
 with petals removed; 6, capsule and bract; 7, petal (minimum size). 
 
MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 67 
 
 Texas. — Wright, without locality, in the Gray Herbarium of Harvard University (?). 
 
 Kentucky. — Neighborhood of Lexington, Short, without date, in the herbaria of Columbia 
 University and Academy of Natural Sciences, Philadelphia. The sheet in the Columbia 
 University herbarium bears the note "Oenothera muricata. Vespertine, 3 to 5 feet high." 
 This is the sheet referred to (MacDougal, Vail, Shull & Small, 1905, p. 6) as belonging to 
 O. lamarckiana, but which in the writer's opinion should belong here. The Philadelphia 
 specimen gives the exact locality "Neighborhood of Lex" fl. Augt. & after, i mile from town 
 left of Coles road to Frankfort. " It may be an escape from cultivation. 
 
 The rediscovery of the original type locaHty of this species is recorded in a 
 previous paper (MacDougal, Vail, Shull & Small, 1905, pp. 7 and 8). The 
 seeds from which the plants described above were raised were sent to the New 
 York Botanical Garden from Tensaw, Alabama, by Mrs. J. F. Davis at the 
 request of Professor Tracy, absolutely ripe capsules not having been found at 
 the time of his visit late in August, 1904. In cultivation, 0. grandiflora has 
 the appearance described by Bartram in 1793 and again by Professor Tracy in 
 1904. (MacDougal, Vail, Shull & Small. 1905, pp. 7 and 8). 
 
 It is very fragrant and showy, and flowers abundantly. Most conspicuous 
 and characteristic are the tapering pale-yellow buds with long, slender, seta- 
 ceous tips to the calyx-lobes. They have none of the heavy, swollen appear- 
 ance of those of O. lamarckiana and of some of its derivatives. For 24 hours 
 or so before the opsning of the flower the closed stigma crowded into the apex 
 of the bud is distinctly seen through the thin epidermis of the closed calyx- 
 lobes. The calyx-lobes when expanded are split open in twos (but very rarely 
 in fours), the petals only attaining their full size during expansion. This is 
 also the case in Oenothera argillicola (Onagra argillicola McKenzie). (Mac- 
 Dougal, Vail, Shull & Small, 1905, p. 12.) 
 
 A further comparison of supposed plates of 0. grandiflora with living speci- 
 mens of the species would indicate the following : The plate of ' ' Oenothera 
 lamarckiana" in Ivcmaire (Illustration horticole, 9, p. 318, 1862) appears to 
 be that species, but the description of the 600 flowers, buds, and capsules on 
 one single plant would appear to refer to 0. grandiflora, 2iS O. lamarckiana does 
 not (in cultivation) flower so abundantly. The plate in Edward's Botanical 
 Register No. 1604 of ''Oenothera biennis var. grandiflora" has much stouter 
 bud-tips, bracts that are broader at the base, and much broader petals than 
 the typical O. grandiflora. 
 
 The plate Oenothera grandiflora Sims (Curtis's Bot. Mag., pi. 2068) is quite 
 typical of the plant grown in the New York Botanical Garden as to the shape 
 of the petals, but the bud-tips are much too heavy and the bracts too broad 
 and clasping at the base. 
 
 Large-flowered evening-primroses have appeared from time to time in the 
 eastern seaboard States; but none that have been examined so far can, be 
 determined as being certainly indigenous there. In the Gray Herbarium the 
 following specimens are noted. 
 
68 MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 
 
 New Jersey. — Pine barrens, Miss Treat of Vineland, 1871, "wild." This has a flower 
 that when expanded must have measured nearly 10 cm. in diameter, the petals being 5 cm. 
 long and over, but the bud is stouter and heavier than 0. grandiflora and has more the char- 
 acter of that of O. lamarckiana. It might be interesting to know whether it still survives in 
 the pine barrens. 
 
 Maine. — Orono. M. L. Fernald, July 17, 1892. A specimen with very much the appear- 
 ance of Oenothera rubrinervis, having the very characteristic bud of that derivative of O. 
 lamarckiana. It is probably not indigenous. Large-flowered specimens labeled Oenothera 
 biennis var. grandiflora and Oenothera grandiflora from California appear to be Oenothera 
 hookeri Torrey & Gray. 
 
 OENOTHERA SIMSIANA SeringE. 
 
 Oenothera corymbosa Sims. Curtis's Bot. Mag., 45: pi. 1974, 1818. Not Lamarck. 
 Oiiagra spectabilis Spach. Nouv. ann. mus. Paris, 4: 352, 1835. (Repr. 32, 1836.) 
 Oenothera simsiana Seringe, in DC. Prodr., 3: 47, 1828. 
 
 Seedling about 6 weeks old. — Leaves glabrate; blades oblong, those of the later leaves oval, 
 20 to 35 mm. long, 10 to 15 mm. wide, obtuse at the apex, gradually tapering into the 
 margined petiole, light yellow-green (fig. 72). 
 
 Seedling about 5 months 
 old. — Rosettes rather loose, 
 40 to 45 cm. in diameter, 
 spreading; outer leaves 15 to 
 20 cm. long, 4 to 6 cm. wide; 
 blades oblong to obovate, 
 broadest somewhat above the 
 middle, taperingto the obtuse 
 apex or in the younger leaves 
 to an acutish apex, gradu- 
 ally tapering into the broad 
 petiole, light yellow-green, 
 obscurely puberulent, with 
 widely scattered hairs on 
 both surfaces (under a lens), 
 approximately repand- 
 denticulate with rather dis- 
 tinct shallow teeth; petiole 
 white, 8 mm. or more wide. 
 
 Mature rosette. — Leaves loosely spreading, finely but obscurely pubescent all over, the 
 larger ones 20 to 28 cm. long, 5 to 6 cm. or more wide; blades oblong to oblong-obovate, 
 shallowly repand-denticulate, tapering into a very wide white-margined petiole (plate 13). 
 
 Adult plant. — Plant rather straggling in appearance, appressed-pubescent and hispidu- 
 lous, especially in the upper portions. Main stem about 2 m. in height, greenish-white, 
 becoming reddish, stout, terete, or slightly angled and sulcate above, irregularly branched 
 at the base, the long, rather weak, virgate branches ascending; leaves minutely pubescent 
 above, somewhat less so beneath, 7 to 10 cm. long, blades obscurely repand-denticulate, 
 oblong-lanceolate to lanceolate, acute or acuminate at the apex, abruptly narrowed into a 
 very short margined petiole or the uppermost sessile, coriaceous, very brittle, bright yellow- 
 green, soon turning red, twisted, the upper ones more markedly so; terminal rosette-like 
 cluster of the inflorescence symmetrical, the floral bracts divaricate and somewhat reflexed; 
 bracts lanceolate, acuminate, subsessile and subcordate; conic portions of the bud 2 cm. 
 long, 6 to 7 mm. in diameter at the base, canescently appressed-pubescent and hirsute, 
 
 KiG. 72. — Oenothera simsianasecdling about 6 weeks after germination. 
 
PLATE 13. 
 
 Oenothera simsiana. 1, leaf from mature rosette; 2, stem-leaf; 3, bud; 4, flower with 
 petals removed and bract; 5, capsule and bract; 6, petal. 
 
PLATE 14. 
 
 t.\ 
 
 ■■:;< 
 
 
 
 
 Adult rosctti.: of" Oenothera simsiana (upper left-hand corner). Mature plant of Oenothera sinisiana in 
 experimental s^rounds in New York, hearing flowers and capsules. 
 
MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 69 
 
 tapering in terminal portion only, the erect, free tips of the sepals, 3 to 4 mm. long, mostly- 
 unequal; hypanthium 4 to 5 cm. long, very slender, densely appressed-pubescent; ovary 
 1 2 to 13 mm. long, slender, densely appressed-pubescent ; sepals 3 to 4 cm. long, shorter than 
 the tubular portion of the hypanthium ; petals thin, bright golden-yellow, fading saffron- 
 yellow, 4 cm. long, 4 to 5 cm. wide, deeply emarginate; filaments 25 mm. long, very slender; 
 anthers slender, 12 to 15 mm. long; pistil as long or slightly longer than the stamens; lobes 
 of the stigma 6 to 7 mm. long, divaricate; capsule about 3 cm. long, 6 to 7 mm. in diameter 
 at the widest portion, 4-angled, appressed-pubescent, rather abruptly contracted at the 
 apex (plate 14). 
 
 Type locality of the plants described above, along International Railroad, 
 City of Mexico. Collected by J. N. Rose and Jos. H. Painter, September 20, 
 1903. No. 7219. 
 
 Seeds of a late-flowering plant were sent to the New York Botanical Garden, 
 unnamed, by Dr. Rose late in 1904, and the plants from which the above 
 description was compiled were grown under glass for the first 6 months and 
 then transferred to the experimental grounds. The flowers produced at the 
 height of maturity are somewhat larger than those received with the seeds, 
 but the late summer flowers exactly resembled those from the original locality. 
 Remarkable for the form and color of the light green rosette and for the very 
 characteristic assurgent or erect habit of the slender virgate stems and the 
 acuminate, curved, and twisted leaves. 
 
 This species is referred to Oenothera simsiana, perhaps somewhat doubt- 
 fully. The plate (1974) in Botanical Magazine agrees with it, except for the 
 corymbose inflorescence, which may occur readily enough in the plant in its 
 native habitat, as many related species var}" in that regard. The mature and 
 normal capsules do not show the strongly reflected, white valve-tips that are 
 seen in the illustration, but the plant as cultivated in the New York Botanical 
 Garden was subject to the sting of an insect that caused a malformation of the 
 ovary or young capsule, which then had very much the appearance of the cap- 
 sules in Sims's plate, though when malformed they never quite reached the 
 state of maturity depicted there. This is what may have occurred at Long- 
 leats in 18 16. 
 
 0. simsiana, with the synonymy as given above, is listed in Hemsley (Biol. 
 Centr. Am., 1 : 454) without any indication of definite locality or collection. 
 
 An unnamed Oenothera collected at the city of Durango and vicinity, by 
 Dr. Edward Palmer (No. 293, 1896, in the herbarium of the New York Botani- 
 cal Garden) can be referred here with a fair degree of certainty. 
 
 O. simsiana is closely related to Oenothera hookeri Torrey & Gray, with 
 which it has doubtless been confounded in herbaria. The growing plants are, 
 however, very dissimilar from that species in general aspect and habit from 
 the rosette stage to maturity, and the flower appears to retain its yellow color 
 very much longer than does 0. hookeri. 
 
70 MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 
 
 OENOTHERA OAKESIANA (Robbins) S. Watson. 
 
 Oenothera biennis var. oakesiana Robbins. A. Gray, Man., ed. 5, 178, 1867. 
 Oenothera oakesiana S. Watson. Biblio. Ind. N. Am. bot., 383, 1878. 
 Onagra oakesiana Britton. Mem. Tor. bot. cl., 5: 233. 1894. 
 
 Seedling about 2 months old. — Leaves obscurely puberulent, with few scattered hairs; 
 blades oblong or oblong-obovate, 6 to 8 mm. v.ide, rounded or obtuse at the apex, tapering 
 to a slender margined petiole (plate 15, a). 
 
 Seedling about 5 months old. — Rosette crowded, symmetrical, 7 to 8 cm. in diameter; outer 
 leaves 3 to 4 cm. long, 8 to 10 mm. wide; blades oblong-ovate, broadest above the middle, 
 obtuse at the apex, tapering to the white, margined, entire petiole, dull or pale blue-green, 
 rather fleshy, approximately denticulate above, glabrous, except for a few obscure hairs on 
 the margins. 
 
 Seedling 8 months old. — Rosette 9 to 10 cm. in diameter, very symmetrical and crowded, 
 raised above the ground with a few of the old leaves adhering to the rootstock; blades approx- 
 imately and more prominently denticulate. 
 
 Mature rosette. — Leaves more or less strigose-pubescent all over, the central ones quite 
 densely so, the larger ones from 15 to 20 cm. long; blades narrowly lanceolate, approxi- 
 matelyand shallowly toothed at the acute apex, more deeply toothed at the slender, taper- 
 ing base (plate 15, b). 
 
 Adult plant. — Plant i to 1.5 m. high, rather slender, clothed nearly throughout with a fine, 
 close, strigose, appressed, and somewhat cinereous pubescence and few scattering longer 
 spreading hairs, becoming almost glabrate and glaucous with age. Stem light or whitish 
 in color, angled and channeled above, branched to the middle; leaves 10 to 15 cm. long; 
 blades shallowly and remotely toothed, often more deeply so at base, narrowly lanceolate, 
 acute and tapering at each end, sessile or nearly so, gray-green and shining above, paler 
 beneath, rather thick and somewhat brittle; terminal rosette-like cluster of the inflorescence 
 symmetrical, the floral bracts very small, divaricately spreading; bracts narrowly lanceolate, 
 more deeply and regularly toothed than the leaves, acuminate at the apex, tapering to the 
 base, sessile, becoming 2.5 times as long as the mature capsule; conic portion of bud 10 to 
 13 mm. long, rather prominently 4-angled, 4 to 5 mm. in diameter at the base, the free 
 separate divaricate tips5mm. long, finely but inconspicuously appressed pubescent; hypan- 
 thium 2.5 mm. to 2.7 mm. long, slender, finely but sparingly pubescent; calyx-lobes half as 
 long as hypanthium; petals firm, 13 to 15 mm. long, 12 to 14 mm. wide, deeply emarginate, 
 not opening widely; filaments 12 mm. long; anthers 7 mm. long; pistil a? long or slightly 
 shorter than the exserted stamens, the lobes erect, 4 to 5 mm. long; capsules 3 to 3.5 mm. 
 or more long, 7 to 9 mm. in diameter, angled and rounded, finely appressed-pubescent, 
 abruptly constricted near the apex. Remarkable for its large seeds (plates t6 and 17). 
 
 Massachusetts. — Uxbridge, J. \V. Robbins, Au.gust, 1878. 
 
 Rhode Island. — Providence, E. P. Bicknell, August, 1896; September, 1899. 
 
 New York. — New York Botanical Garden, D. T. MacDougal, September, 1904; Cold 
 Spring Harbor, Long Island, G. H. Shull, September, 1904. 
 
 The above specimens are the only ones in the herbaria of the New York 
 Botanical Garden and Columbia University that can with any degree of 
 certainty be referrred to 0. oakesiana. The Robbins specimen is from the 
 herbarium of the late Rev. Thomas Morong and attached to its sheet is the 
 following note in the writing of its discoverer. 
 
PLATE 15. 
 
 B 
 
 A. SeL-dliiiij of Oenotlifni (>:ikt;sian:i about two inoiitlis old 
 
 B. Ailult rosette of Oenothera oakesiana. 
 
PLATE 16. 
 
 Oenothera oakesiana. l.leaf from mature rosette; 2, stem-leaf; 3, capsule and bract; 
 4, flower with petals removed; 5, buds; 6, petal (minimum size). 
 
PLATE 17. 
 
 ()eii(>tlic'i:i oake^iaiKi "rown wild in waste land in New \'i)rk Botanical (iardcn. 
 
MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 7 1 
 
 Oenothera oakesiana mihi. 
 
 This plant was presented to Professor Gray many years ago as a well-characterized and 
 distinct species, which I still consider it. I discovered it in the medical colleges in New 
 Havea, Connecticut, in 1S27. In 1831 1 sowed its seed with those of the ordinary variety. 
 The differences are as follows: 
 
 (i) If raised from seeds sown in the spring it is annual, the common variety by its 
 side being biennial. 
 
 (2) The leaves of the first year, when biennial, are much narrower. 
 
 (3) The pubescence throughout is soft-appressed, that of the other being coarser and 
 spreading. 
 
 (4) The points of the calyx are always spreading, in the other appressed. 
 
 (5) The capsule is longer and more taper. 
 
 (6) The ripe seeds are larger. 
 
 I have found but two other localities, namely, on sides of the railroad north of Norton, 
 Massachusetts, and another, same situation near Apponong Depot, Rhode Island. Also it 
 is well naturalized in Uxbridge, Massachusetts, from my seeds. 
 
 UXBRIDGE, 5th Sept., 1878. J. W. ROBBINS. 
 
 The plants from the New York Botanical Garden were first noticed by Dr. 
 MacDougal late in the summer of 1904. They grew not far from the embank- 
 ment of the Harlem Division of the New York Central Railroad and were 
 quite abundant, though not covering a very widespread area. With them 
 were associated a number of 0. biennis, the latter in that particular spot, how- 
 ever, being not as abundant as elsewhere in the garden. Rosettes were 
 obtained of the O. oakesiana and the plants raised from them the following 
 summer were in every way similar to the wild plants, and formed the basis of 
 these notes on the species. Plate 17 was photographed from a specimen 
 growing wild in the New York Botanical Garden. 
 
 OENOTHERA PARVIFLORA Linnaeus. 
 
 Oenotliera parviflora Linnaeus. Syst., ed. 10: 998, 1759. 
 Onagra parviflora Moench. Meth. Supp!., 287, 1802. 
 
 Seedling about 6 weeks old. — Rosette 2.5 to 3 cm. in diameter, rather loose and irregular, 
 blades of the leaves ovate or ovate-spatulate, glabrate, ciliate. 
 
 Rosette about 3 months old. — Leaves finely pubescent on both surfaces, less so above; 
 blades various, those of the earlier leaves narrowly oblong or oblong-spatulate, varying to 
 oblong-obovate, 6 to 10 cm. long, the larger ones 2.5 to 3 cm. wide, approximately and 
 shallowly denticulate, obtuse or nearly so at the apex, gradually narrowed to the base of the 
 margined petiole, which is very early streaked with red or pink. 
 
 Mature rosette. — Leaves long and slender, spreading flat on the ground; blades lanceolate, 
 or oblong-lanceolate, acutish, irregularly and strongly denticulate, more deeply toothed at 
 the long, tapering base, dark green and shiny, sparingly mottled with red, sparingly puber- 
 ulent above, more so and paler beneath, 15 to 30 cm. long, 3 to 5 cm. wide; petiole white, 
 8 mm. wide, margined to base, distinctly pink or red ringed most of its length (plates 18 
 and 20). 
 
 Adult plant. — Central stem not always maturing the first season when forced. Lateral 
 branches of annual rosette stout and angled, pubescent with spreading hairs; leaves 
 resembling those of the rosette in color, texture, and pubescence; blades oblong to oblong- 
 
72 MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 
 
 lanceolate, tapering to the short, margined petiole, shallowly and approximately denticulate 
 above, some of them crinkled close to the midvein; terminal rosette-like cluster of the inflo- 
 rescence symmetrical and spreading; bracts lanceolate, acute, sessile or nearly so, divaricate, 
 crowded on short internodes; flowers numerous; bud club-shaped, the conic jjortion 8 mm. 
 long, 3 to 4 mm. in diameter, rarely angled, the free tips separate, quite slender, erect, 2 nmi. 
 long or less, mostly tipped with red: hypanthium 3 to 3.5 mm. long, slender, i)uberulent; 
 ovary 8 to 9 mm. long; calyx lobes one-third as long as the hypanthium; petals firm, S mm. 
 long, 9 mm. wide, deeply emarginate, cuneate; filaments 7 mm. long; anthers 4 to 5 mm. 
 long, pistil shorter than the slightly exserted stamens, the lobes divaricate, 3 to 4 m.m. long, 
 capsules abundant, 2.5 cm. long, 6 mm. in diameter at the widest portion, bluntly 4-angled, 
 pubescent, bright green, much thinner in apical portion. (The apex of some of the valves 
 are occasionally more or less distinctly bifid.) (Plate 19.) 
 Illustratioti. — Meerburg, Plantas rariores, pi. 34, 1789. 
 
 Early in 1906 a package of seeds distributed as Oenothera parviflora was 
 received from the Botanic Garden of Madrid, Spain. The plants raised from 
 these seeds differed essentially from any of the various so-called O. biennis of 
 European origin in cultivation in the experimental grounds of the New York 
 Botanical Garden. The plants were very persistently biennial, and only one 
 or two of those placed in the open threw out lateral branches, the main branch 
 only starting its growth too late in the autumn to mature. Specimens of any 
 Oenothera closely resembling it are not to be found in the herbarium of the 
 garden or in that of Columbia University. As 0. parviflora Linnaeus was 
 credited to "Canada to Virginia" by Pursh (Fl. Am., Sept., 261, 1814), with the 
 further note that it was rare, the development of the Madrid plants was fol- 
 lowed and recorded with the hope that eventually the species might still be 
 found a component part of the North American flora. 
 
 Late in the fall of 1905, Miss N. M. Stevens sent in some rosettes of what she 
 considered as two distinct species of Oenothera collected in the neighborhood of 
 South Harpswell, Maine. The rosettes were placed in the experiment? 1 
 greenhouse for the winter and later Miss Stevens sent seeds of both species 
 from the same locality. These latter germinated in the greenhouse and 
 though somewhat similar in the early rosette stage, one of them, the "pre- 
 vailing species" of the locality from whence the seeds came, was identified 
 with the plant determined as 0. muricata Linnaeus and the other, the "rarer 
 species" growing on sandy wastes, was identical with the species grown from 
 the Madrid garden seeds under the name of 0. parviflora, and it is from the 
 plants raised from Miss Stevens's seeds that this description has been drawn 
 (plate 20). 
 
 The species shows distinct characteristics in color and habit. The flowers 
 are but rarely exserted beyond the bracts, which are so crowded on short 
 internodes that the ends of the branches have a closely-tufted appearance. 
 
 The Linnaean description under Oenothera reads: "Parviflor. A. OE. fol. 
 ovate-lanceolatis planis, caule laevi subvelloso. Margo coronans jructum, non 
 uti praecedentis quadrifldus , sed octofidus." 
 
PLATE 18. 
 
 \.A 
 
 Ocnotlicni p;n-viriora gn)\vn from seeds obtained from the notanical Garden of Madrid. 
 
FL4.TE 19. 
 
 2 1 
 
 Oenothera paniflora. 1, seedling two months after germination; 2, leaf of young rosette; 
 3. stem-leaf; 4, bud and bract; 5, flower with petals removed; 6, flower and 
 bract; 7, petal; 8, capsule and full-grown bract (from seeds from Maine). 
 
PLATE 20. 
 
 *i^•^-*- 
 
 •f 
 
 (Xiiotlicra p:ii\ itlora grown from scuils ot plants native near Soulli Ilarpswell, Maine. 
 
PLATE 21. 
 
 Oenothera muricata. 1, leaf from mature rosette; 2, stem-leaf; 3 and 4, young bracts and 
 buds; 5, flower with petals of maximum size; 6, petal; 7, flower with petals 
 removed; 8, capsule with bract; 9, capsule. 
 
MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE) OENOTHERAS. 7J 
 
 The last character would appear to be an accidental one, that is, if the name 
 belongs at all to the plant in question here, and it is extremely doubtful if 
 that point can ever be definitely settled. 
 
 Some of the capsules of the Madrid plant showed a distinct bifid termination 
 of the valves of the capsules, but in the Maine plant that character was less dis- 
 tinct or altogether lacking. The Spanish plants and those from Maine, grow- 
 ing side by side, were not distinguishable, but the Maine plants matured more 
 rapidly. 
 
 OENOTHERA MURICATA Linn^us. 
 
 Oenothera muricaia Linnaeus. Syst , ed. 12, 26.^, 1767. 
 Onagra muricaia Moench. Metli., 675, 1794. 
 
 Seedling about 6 weeks old. — Rosette crowded and a little irregular, 2 to 3 cm. in diameter, 
 blades of the leaves oblong, glabrate, slightly ciliate on the margins. 
 
 Seedling about 3 months old. — Rosettes crowded, 7 to 8 cm. in diameter; leaves 3 to 3.5 cm. 
 long, 7 to 8 mm. wide, blades oblong-spatulate, broadest above the middle, obtuse at the 
 apex, tapering into the margined petioles, blue-green, fleshy, glabrate or becoming appressed- 
 pubescent as the leaves increase in size, obscurely denticulate, the apex of the denticulations 
 mostly reddish. 
 
 Mature rosettes about 5 months old. — Rosetts 15 to 17 cm. in 
 diameter, rootstock thick, raised 2 to 3 cm. above the ground, 
 bearing a few old leaves under the later green ones; leaves 
 crowded, the outer 9 to 14 cm. long, 1.5 to 2 cm. wide; blades 
 approximately denticulate, irregularly and more deeply so towards 
 the base above the broad, white-margined petiole, appressed- 
 pubescent or hirsute, v>ithout any trace of red in the leaves, 
 except at the apex of the denticulations. 
 
 Adult plant. — Plant 1.5 m. high, pubescent with short appressed 
 as well as longer spreading hairs throughout. Stems erect, stout, 
 
 1 to 1.5 m. high or more, single or generally branched above, 
 angled, turning red and the outer bark splitting into shreds at 
 maturity; stem-leaves rather crowded, loto 15cm. long, 1.5 to 
 
 2 cm. wide; blades narrowly lanceolate or linear-oblong, acute, 
 approximately denticulate, more deeplyso atthebase, appressed- 
 
 pubescent with short strigose hairs more or less on each surface, ^^^- 73.— Seedling of Oeno- 
 
 ,., ., ,.,, J., Ihera muricaia about 2 
 
 light green, paler beneath; termmal rosette-like cluster or the months old. 
 inflorescence barely symmetrical; midvein broad, whitish; inflo- 
 rescence crowded; bracts lanceolate, acute at the apex, tapering to the sessile base, 
 finally two or more times as long as the mature capsule; conic portion of bud 1 1 to 12 mm. 
 long, 4 mm. in diameter, pubescent with appressed and spreading hairs, the free tips 4 mm. 
 long; hypanthium 2.5 to 3 cm. long, pubescent; calyx-segments about half as long as the 
 hypanthium; ovary i cm. long; petals thick, 10 to 12 mm. long, 11 to 14 mm. wide, deeply 
 emarginate, not opening widely; filaments about i cm. long; anthers 4 to 5 mm. long; pistil 
 shorter than stamens; stigmatic lobes erect, thickish, 3 to 4 mm. long (plate 21). 
 
 This description is based on plants raised from seeds secured from Long 
 Island (Bicknell, 1904, and Vail, 1904) and grown two successive years in the 
 New York Botanical Garden. 
 
74 MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 
 
 Tha Linnsean description reads as follows: "OE. fol. lanceolatis planis, 
 caule purpurascente muricato. Similis parviflorae, sed Fructus os non S-fidis. 
 Caulis puncta rubra sparsa. Canada." (L. Syst., ed. 12, 263. 1767.) 
 
 Based on this description and without reference to any later interpretation 
 of the species, the following specimens are placed here : 
 
 Maine. — York, Bicknell, 1896; Cape Neddick, Bickiiell, 1896; South Harpswell, Stevens, 
 1905. 
 
 Massachusetts. — Ipswich, Morong, 1872; Nantucket, Bicknell, 1899. 
 
 New York. — Long Island, Edgemere, Bicknell, 1902 and 1904; Coney Island, Vail, 1904; 
 Staten Island, New Dorp, Kearney, 1894. 
 
 0. muricata L. raised from seed received from Professor De Vries from the 
 Holland sand-dunes resembled these American plants, but were not abso- 
 lutelv identical. This should also be said for some plants that were raised 
 from seed collected by Professor De Vries near Chicago, Illinois, in 1904. 
 Undoubtedly 0. muricata has a wider distribution in the northeastern States 
 than can be noted here. 
 
 In addition to these quite a number of plants that apparently are referable 
 to this species are preserved in the herbaria of the New York Botanical 
 Garden and of Columbia University. They are as follows: 
 
 Canada. — Province of Quebec, New Carlisle, Williams & Fernald, 1902; Rimouski 
 County, Collins & Fernald, 1904; River Ste. Anne des Monts, Collins & Fernald, 1905; 
 Anticosti, Jupiter River, Macoun, 1 883. 
 
 Massachusetts. — Provincetown, Hollick, 1901; Nantucket, MacDougal, 1905. 
 
 Rhode Island. — Block Island, Hollick, 1897. 
 
 The Nantucket, Provincetown, and Block Island plants are probably 
 stunted specimens merely, but the Canada specimens are quite remarkable 
 in that the mature plant still preserved the remains of the rosette of the pre- 
 vious year at the base of the stem and in general appearance ; even to the bent 
 habit of the upper portion of the stem they can be easily identified with the 
 plate and description of 0. muricata in Flora Danica, pi. 1757, 1823. 
 
 An attempt was made to raise some plants from seed taken from Messrs. 
 Collins & Fernald's specimen from River Ste. Anne des Monts, 1905, but it 
 was not very successful. The plants were quite persistently biennial and in no 
 case was a normal central stem secured, and the rosettes in the garden were 
 very much larger than those on the herbarium specimens. The Canada plants 
 are evidentlv a more northern form of the Long Island plant (plate 22). 
 
 The habit of the rosette of the previous year remaining at the base of the 
 flowering stem of the second year's growth is one that is also claimed for his 
 species 0. ammophila by O. Focke (1904). 
 
 The plate of 0. muricata published by Murray (Nov. comm. soc. reg. Sci. 
 Goett., 6: 24, pi. i, 1776) appears to have broader leaves than the American 
 plants enumerated here. 
 
PLATE 22. 
 
 Ounolbcra inuiicata grown iR-;ir the- nortlicrn limits of the species in (^lebcc. 
 
MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 75 
 
 GENERAL DISCUSSION. 
 
 HISTORICAL CONSIDERATIONS. 
 
 With the increasing attention being enhsted in the subject of mutations in 
 organisms, numerous records of the saltatory activation and latency of char- 
 acters in lines of descent are being disclosed by workers in all branches of bio- 
 logical science. Some observations published by Dr. Arthur Hollick in March, 
 1879 (Holhck, 1879) are among the most suggestive. The views expressed by 
 Dr. Hollick over 26 years ago harmonize so well with the sum of informa- 
 tion available at the present time that it will be profitable to give them in 
 full in this place. He saj^s: 
 
 The object is more to call attention to a few facts which seem to have been generally 
 passed over in botanical researches as devoid of interest. The whole subject arranges 
 itself under two heads. The first will comprise true "albinos" or such plants as have 
 spontaneously, in a state of nature, lost their colors and become white flowered. The second 
 relates to those plants in which the colors have been more or less eliminated by artificial 
 means. 
 
 First, then, we have to consider those "sports" of nature where there has been a sudden 
 change, without any intermediate steps, from a plant with colored flowers to a pure white 
 variety; which change, for want of a better term, we may call "spontaneous. " Such may be 
 aptly termed "negative" varieties, since their pecularity is due rather to an absence of their 
 normal color than to the presence of white. 
 
 It is nothing uncommon to see, in many species, a gradual change from a brightly colored 
 individual, through successive lighter and lighter ones, until a pure white is reached. This 
 may be very well seen in Hepatica triloba, which comprises individuals of every shade from 
 dark purple to white; or in Polygala sanguinea, in which we find the same gradual change 
 from a dark red. * * * 
 
 The following list of "albinos" is made from specimens collected during the last three 
 years. Vemonia novaboracensis W'illd., Lappa officinalis var. major, Lobelia syphiliiica 
 Linn., Epiphegus virginiana, Verbena hastata Linn , Asclepias incarnata Linn., TrifoHum 
 pratense Linn., and Brunella vulgaris Linn. Both Gentiana crinifa Froel. and Lobelia 
 cardinalis Linn, have been reported to nie as having produced at times albino forms, but I 
 have never seen them personally. * * * 
 
 Now, in the first place, not only does the flower show the characteristic absence of color 
 but the leaves, stem, and, in fact, the entire plant, are invariably of a lighter green; and if 
 any red or green should be normal to the stem (which is often the case) this will also be of a 
 lighter shade. * * * 
 
 Secondly, if we have under consideration a plant which commonly is known to have 
 juice of an acid or peculiar taste, this is generally more or less absent in the albino form, and 
 sometimes is eliminated entirely. Darwin has noted this fact, and, in commenting upon it, 
 says that honey bees evidently are aware of it, for they perforate the calyx and corolla of 
 the white Aconitum napcllus, to get at the nectaries, but will not so do with the colored 
 ones. * * * 
 
 It has often been urged that these albinos are mere "sports" of Nature, with nothing 
 constant about them; their peculiarities due, often, to growing in the shade, etc. In fact, 
 that it is a condition due to bleaching or insufficient sunlight, and that there is nothing 
 
76 MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 
 
 inherent in the constitution of the plant. Fortunately I have been able to test this. In 
 the case of Lobelia syphilitica I first found the plants in the shade of some rather thick 
 underbrush, in the month of September. This growth was cut down the following spring, 
 and the place opened to the full glare of the sun. This was done three years ago, yet, every 
 autumn since, the plants have either reappeared, or else left offspring which have inherited 
 their albino nature. This shows them not only constant in their peculiarities, but also that 
 these are bred in the plant and capable of inheritance. Epiphcgus virginiana and Brunella 
 vulgaris offer the same proofs. Nor has the influence of locality much or anything to do 
 with it, for a plant of the Lobelia syphilitica (with white flowers), which originally grew in a 
 swamp, was transplanted to a dry garden a mile or more away, yet came up and blossomed 
 white the next year. 
 
 Let us now see what the experience gained in cultivation of white varieties can tell us. 
 Perhaps the Japanese have brought the art of eliminating color from plants to the greatest 
 perfection. Scores of species and genera have been by them variegated in the most peculiar 
 manner. But this is never constant in this country, but after a while always tends to 
 revert to the primitive color again. * * * 
 
 As there seems to be no tendency to reversion in these natural albinos, they might per- 
 haps be made permanent varieties and be valuable on that account. No doubt this per- 
 manence is due to the change being sudden, leaving no trace of color by intermediate steps, 
 while in cultivated examples the white has generally been obtained by a gradual selection of 
 less and less darkly colored ones, and hence there would be a greater tendency to reversion 
 back through these steps again. 
 
 From the foregoing it is clear that retrogressive white-flowered derivatives 
 were seen to arise from colored parental forms by saltations in which the color 
 of the flower behaved as a unit-character, and furthermore that the new color- 
 less forms were constant and were not affected by environmental conditions. 
 These ideas in fact underlie an important part of the mutation theory, and the 
 statements by Dr. Hollick constitute the unexpanded thesis of the saltatory 
 behavior of one set of characters. Not the least interesting part of the dis- 
 cussion is that which in effect recognizes the fact that many white-flowered 
 and white-leaved forms in cultivation represent the extremes of fluctuating 
 variations, and that when artificial selection lapses these forms gradually 
 return to a normal mean ; also that some species, such as Hcpatica triloba, may 
 also have an extremely wide range of fltictuating variability as to the flower- 
 colors. 
 
 A dim realization of the importance of sports and of the necessity for some 
 other hypothesis besides natural selection to account for the existence of all 
 living plants seemed to be widespread among botanists about the time De Vries 
 began the investigations which have led to his conclusions as to the influence 
 of mutations. An interesting forecast as to the matter was made by the late 
 Thomas Meehan (Meehan, 1882). He said: 
 
 The conclusion I have been forced to is that the odd forms we often find in nature are not 
 necessarily hybrids, but are as likely, if not more likely, to be the outgrowth of some internal 
 law of form with which we are as yet unacquainted. That they do not often perpetuate 
 themselves is \not^ remarkable when we remember that of thousands of seeds produced on 
 anyone tree but a small percentage ever gets a chance to form, and of those which do sprout, 
 
MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 
 
 / / 
 
 again but a small percentage survives to become bearing trees. As the number of trees 
 reproducing the general features of the original may be a hundred to one of the more strik- 
 inglyaberrant forms, we may see that though individual instances may be common, we are 
 never likely to meet many trees of one stamp. Once in a while an individual tree may find 
 itself in a situation favorable to the preservation of a number of seedlings, which might 
 endure until again reproductive; in such cases a marked variety may originate and make 
 its way over the earth. 
 
 I have often thought it probable that in time a few individuals of these suddenly intro- 
 duced forms might again leap into new features, and then if they should be able to sustain 
 themselves we should have new species quite independently of any principle of natural 
 selection. 
 
 The negative implied by the context is supphed in brackets and indicated 
 by italics. 
 
 The above are by no means to be the earliest formulations of the hypothet- 
 ical conclusion as to the importance of sports in phylogeny. Among others 
 which might be mentioned the statement by Kerner in 1869 is worthy of 
 citation. He says: 
 
 Die Arten die uns gegeniibertreten, sind nur Stadien and haben als solclie zwar fiir eine 
 gewisse Zeit Konstanz, konnen sich aber friiher oder spater in andersgeformte Arten auflo- 
 sen. Thatsache ist es, dass alle Pflanzen friiher oder spater einmal vereinzelte Abarten, 
 d. i. S])r6sslinge zu erzeugen in Stande sind, welche in ihren Merkmalen von der Mutterart 
 abweichen, und zahlreichen Erscheinungen drangen uns zu der Annahme, dass unter dem 
 zusammentreffen giinstiger Bedingungen solche Abarten die Ausgangspunkte neuer Arten 
 werden. 
 
 Among other recent observations is that of Schaflfner (1906) concerning 
 a derivative of Verbena stricta which bore pinkish-white instead of the deep 
 purple flowers of the species. The aberrant form was present in many thou- 
 sands of individuals among the typical specimens, with which it seemed to be 
 in successful competition. Tests have not yet been made to determine 
 whether or not it does not readily cross with the parental form, or whether it 
 is dominant or recessive with regard to the more important differentiating 
 character. In any case, however, its mutative origin is based upon evidence 
 similar to that which zoologists seem willing to accept as proof of mutation in 
 animals, and its successful maintenance against parental competition is an 
 established fact. 
 
 Many of the observations brought forward by various authors concern 
 species which have been long under cultivation, and so many facts reported 
 from horticultural material have been found to be unguarded against errors 
 affecting purity of lineage that much prejudice exists with respect to data 
 obtained from domesticated forms. So far as the effects of actual tillage or 
 domestication are concerned, these prejudices are without foundation, since 
 at the present time no well-founded evidence exists to show that cultural con- 
 ditions have ever produced alterations which were strictly and continuously 
 inheritable under environic conditions other than those by which the altera- 
 
7$ MUTATIONS, VARIATIONS, AND RELATIONSHIPS OP THE OENOTHERAS. 
 
 tions were induced, although vague statements and erroneous generahzations 
 to the contrary are current. The possibiHty of such results is to be by no 
 means denied, but if this view is to be upheld it must be supported by some 
 evidence based upon actual experimentation. Vicinism, the somatic multipli- 
 cation of bud-sports and extreme variants, and the confusion of closely related 
 elementary species form the basis for the greater number of positive assertions 
 as to the effects of cultivation, and it is necessary to examine all facts bearing 
 upon the lineage of supposedly new forms with the greatest care before their 
 aspect or behavior may be taken as evidence upon phylogenetic problems. 
 
 An extended consideration of the phenomena of mutability has led the 
 authors of this paper to entertain grave doubts as to the practical value of 
 the conception of periodicity. De Vries throughout his writings has favored 
 the assumption, on purely theoretical grounds, that a species must, at certain 
 times in its history, be in a mutable condition, while in other periods of great 
 length all of the progeny come true to the type, subject only to fluctuating 
 variability. 
 
 It is quite possible that any given form may give off derivatives steadily 
 for centuries, but none of these being fitted to survive, no trace of this activity 
 remains. A change in the character of the mutants originated, or newly 
 found conditions met with in comparatively short migrations, might readily 
 give an appearance of the beginning of a mutative period, unless the matter 
 had been tested by pedigree-cultures. 
 
 With regard to the actual succession of generations two aspects of the case 
 present themselves. In one case it may be supposed that an annual species 
 might have a limited distribution, and, occupying small areas, might be repre- 
 sented by only a few individuals every year. Combined with these circum- 
 stances, the proportion of mutants produced by the species might be so small 
 that it would be possible to examine every specimen for a hundred years 
 without meeting a single atypic individual. The conclusion would then be 
 reached that this was included in the immutable period. Then if the number 
 of individuals grown and examined in the next ten years numbered as many 
 millions, and some mutants were found, the line of reasoning followed would 
 lead to the fallacious assumption that a mutable period had begun, or rather 
 that some physiological change had ensued in consequence of which mutations 
 were occurring. These might as readily be seen during the first year of 
 observation, however, if enough individuals had been grown. 
 
 It appears, therefore, that the real state of affairs is better represented by 
 the phrase "frequency of mutation," by which is expressed the number of 
 individuals which must be grown to furnish one mutant, and which is nearly 
 identical with "the coefficient" of mutability. If the parental type pro- 
 duces more than one mutant, the frequency of these may vary widely from 
 each other. 
 
MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 79 
 
 Thus of the great number of mutant individuals originated by 0. lamarck- 
 iana but few have been ruhrinervis, still a smaller number represented gigas, 
 while the combined observations at New York and Amsterdam have failed to 
 bring hrevistylis within the range of the cultures, this form having arisen at 
 Hilversum, where it is still in existence. To secure mutants, or to allow a 
 species to display its possible derivatives, the long stretch of centuries may be 
 shortened and as many individuals examined in a seedling stage in a single 
 season as might naturally mature in twenty decades. The real point to be 
 attained is to secure sufhcient individuals to include the proportionate num- 
 ber of all the derivatives. 
 
 ADDITIONAL EVIDENCE AS TO THE DISTRIBUTION AND OCCURRENCE 
 
 OF LAMARCK'S EVENING-PRIMROSE. 
 
 The records and material cited and examined show that a large-flowered 
 evening-primrose, which can be identified with nothing so closely as with 
 0. lamarckiana, had found its way into the gardens near Haarlem, Holland, as 
 early as 1756, several years before 0. grandiflora was brought from its habitat 
 in Alabama to England in 1778. Also that an evening-primrose with large 
 flowers was seen on the sand banks north of Liverpool, England, in 1806, on 
 the coast of Somerset in 1837, and has been cited and described by various 
 authors in 1845, 1851, 1855, i860, and continuously since 1892. Plants 
 grown from seeds obtained from the region in question proved to be 0. lamarck- 
 iana with two of its mutants, rubrincrvis and lata. Not only did this con- 
 firm the supposition as to the long history of lamarckiana, but it also showed 
 that rubrincrvis had successfully maintained itself in competition with the 
 parental form, while lata had become capable of self-fertilization, the only 
 instance on record. 
 
 THE MUTANTS AND MUTABILITY OF LAMARCK'S EVENING-PRIMROSE. 
 
 The observations begun by De Vries upon 0. lamarckiana in 1884 and 
 continued upon purely pedigreed material by himself until the present time, 
 and the cultural experiments described in this paper, demonstrated a very 
 high frequency of mutation in this species and also that the mutative depart- 
 ures show a great diversity when compared with other forms. In addition to 
 the derivatives which constitute the greater proportion of the atypic offspring 
 gigas has recently reappeared in an example in New York, being one of the 
 rarer forms. Then the sudden appearance of a totally new type among the 
 mutants leads at once to the confirmation of the suggestion that certain 
 mutants may appear so infrequently that they might not be expected more 
 than once in a million individuals and consequently would not be seen except 
 at long intervals. 
 
 Of the known mutants alhida, oblonga, gigas, nanella, lata, and scintillans 
 were found in the pedigreed cultures described in this paper, and with greatly 
 
8o MUTATIONS, VARIATIONS, AND RELATIONSHIPS OP THEl OE;NOTHE;rAS. 
 
 varying frequency. In one culture the coefficient of mutability of oblonga 
 was pushed slightly beyond the limit hitherto assigned to it by De Vries and 
 it was found to constitute 54 per cent of the atypic derivatives. A survey of 
 all of the results obtained in the New York Botanical Garden since 1902 does 
 not justify the assertion, however, that any change in the frequency of mutation 
 has been induced by cultural conditions, and the only really new feature 
 noticed consisted in the appearance of a new mutant, which, however, was so 
 unsuited to the climatic conditions that it perished before reaching maturity, 
 an example of the eliminative action of selection in the determination of what 
 mutants shall survive and what shall perish. 
 
 Cultures made from seeds obtained from merchants did not show a fre- 
 quency of the mutants above 1.5 per cent in any instance, which is much less 
 than that observed in the strain growing at Hilversum. The sources from 
 which the seeds were obtained indicate that the species is in a state of muta- 
 bility in various parts of the world, and that it has probably been so for some 
 time. The low frequency of mutation may have been due in part to unsuit- 
 able methods of cultivation, in consequence of which the parental individuals 
 were badly nourished. 
 
 0. hrevistylis was brought to maturity and found to agree in all characters 
 with specimens grown in Amsterdam. This form differs from the parental 
 type chiefly by the retrogressive character of a shortened style, and by gen- 
 eral observation this organ shows a wider range of fluctuation about its average 
 than the same organ in the parent, although no exact examination was made. 
 Despite the fact that this type produces comparatively few seeds it was found 
 by De Vries among the individuals of the parental type in the original locality 
 in which mutation in lamarckiana was observed in 1886, and it has since held 
 its own in competition with the parental type, in a manner demonstrative of 
 the fact that it is possible for a mutant to survive in competition with existing 
 forms. 
 
 O. lata grown from seeds obtained from Professor De Vries and from indi- 
 viduals arising in the pedigreed strains in New York was seen to remain con- 
 stant to its characters. A few tests had been made both in New York and 
 Amsterdam, but no fertilization was secured with its own pollen, although some 
 of it appeared structurally perfect in a microscopic examination. 0. lata 
 appearing in a culture from seeds of 0. lamarckiana from England, however, 
 included some individuals which likewise produced pollen, and also showed 
 the hitherto unknown capacity of being fertilized by it. The seeds obtained 
 in this way gave rise to a progeny which showed only the constituents usually 
 found in a progeny of this plant when fertilized by lamarckiana. 
 
 O. scintillans grown directly from its own seeds, and also as it appeared as a 
 new mutant in the cultures, was constant in its ever-sporting character. Seeds 
 of this species purely fertilized gave rise, in addition to the type, to a few 
 
MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE) OENOTHERAS. 8 1 
 
 individuals of oblonga and lamarckiana, in addition to an unrecognizable form, 
 the representative of which did not reach maturity. Purely fertilized seeds of 
 this form have also been seen to give rise to lata and nanclla. 
 
 O. oblonga was grown directly from its own seeds and also appeared in all 
 of the cultures of the parental type which were spread sufficiently to include 
 its coefficient of mutability. It appears to be constant in its characters and 
 is strikingly differentiated at a very early stage of its development. 
 
 O. albida was also grown directly from its own seed and appeared in the 
 pedigreed cultures of lamarckiana. It appeared to coincide in all of its 
 characters with the descriptions given by De Vries. 
 
 The single test which was made of the effect of crossing the parental type, 
 lamarckiana, with one of its mutants, rubrinervis, gave results in accordance 
 with the conclusions of De Vries in this matter. Rubrinervis is a mutant of 
 supposedly low frequency of occurrence, since it has not been found in any 
 pedigreed culture for several years. It may be regarded as a progressive 
 derivative of the parental type and presumably carries all of the lamarckiaim 
 qualities in a latent condition. On the other hand, the strain of lamarckiana 
 with which it was crossed, and from which it was derived several generations 
 back, must bear the characters which gave rise to this form as a mutant. The 
 two plants must therefore carry the same characters, but some are latent in 
 one and some in the other. 
 
 The origin of species by hybridization, which has been a well -recognized 
 phenomenon for half a century, receives further exemplifications in crosses 
 between 0. cruciata and 0. lamarckiana. The fixed forms which may be pro- 
 duced by this cross have been cultivated in European gardens for some time 
 under the names of 0. cruciata, 0. cruciata varia, and 0. biennis cruciata. The 
 list of species which have originated by hybridization is a long one and need 
 not be recounted here. Singularly enough, this fact seems to be unknown to 
 workers who carry on extensive operations in plant breeding. In the instances 
 cited in this paragraph, the most remarkable feature is that a fixed form or 
 new species may be produced by the hybridization of two parental types of 
 comparatively high frequency of mutability. It seems also pertinent to again 
 sound a warning against the continued practice of taxonomists of ascribing 
 hybrid origin to species, with characters apparently intermediate between two 
 neighboring and known forms. Numerous hybrids of known origin in the 
 New York Botanical Garden have been submitted to competent taxonomists, 
 and in no single instance has a correct diagnosis of the parentage been made. 
 
 The fertilization of the eggs of lamarckiana from the pollen of rubrinervis 
 resulted in a single observation in a progeny which contained 71 per cent of 
 the latter, or mutant, type. This is not a universal result, however, since the 
 proportion has been found to vary from 19 to 74 per cent. The pollination of 
 0. biennis hy the mutant described in this paper gave a progeny in the first 
 
82 MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE) OENOTHERAS. 
 
 generation which conformed to the biennis type. The same result is true of 
 the reciprocal cross. It remains to be seen whether the mutant characters are 
 merely recessive in this combination, or whether a condition of latency has 
 been assumed in which the mutant appears only in its customary frequency. 
 
 The evening-primroses, then, afford the following exact evidence as to the 
 effects of intercrossing : Many of the mutants when crossed with the parental 
 form give progeny composed of the parental form and the mutant in the first 
 generation. In the case of gigas and rubrinervis the proportion of the mutant 
 in the progeny is generally much higher than the parental type. Rubrinervis 
 when grown in close contiguity to the parental form showed only a small pro- 
 portion of crosses, and the newly discovered mutant of 0. biennis also gave 
 but a small proportion of crosses when exposed to danger of hybridization 
 with the parent. Lastly it is to be pointed out that 0. brevistylis, which pro- 
 duces but few seeds, and which is recessive when crossed withO. lamarckiana, 
 has been in continuous existence for 20 years, perhaps much longer, at Hil- 
 versum, where it is in direct competition and danger of being hybridized with 
 the parental form. Experiences in the guarded cultures seem therefore to be 
 confirmed by an analysis of natural conditions. Swamping of new forms, by 
 intercrossing, is a specter which looms vast and shadowy across the visions of 
 writers of a speculative habit, but at the present moment it is believed that no 
 evidence carefully tested by modem methods is available to justify the hypoth- 
 esis, so far as plants are concerned. 
 
 The demonstration of the greater variability of the mutants as compared 
 with the parent form seems to be complete, and this relation is one of consid- 
 erable practical and theoretical importance. The greater variability of phylo- 
 genetically new characters as compared with older ones is not wholly new and 
 unique, though we believe our studies in 1904 gave the first statistical support 
 to the thesis when stated in this form. Dar\Ndn's (1859, Ch. II) view that 
 characters which distinguish species are more variable than those that sepa- 
 rate genera, and that varietal are more variable than specific characteristics, 
 gives the first recognition of this principle, if we take into account that accord- 
 ing to his conception of the relation of these several form-groups the variety is 
 an incipient species and the species an incipient genus. Field (1898) worked 
 out the relation between a varietal and a specific characteristic in the wings 
 of a lepidopterous insect, giving a confirmation of Darwin's rule, but he did 
 not relate his facts in any way to the relative age of the characters studied. 
 It is probable of course that the varietal characteristic was a more recent 
 acquisition than the specific, and this would make his results accord well with 
 the observations of the authors on Oenothera. 
 
 To the student of evolution who wishes to institute pedigree-cultures, the 
 relation here demonstrated should prove of much practical value, for it matters 
 not whether he desires to study the possibility of fixing extreme fluctuations, 
 
MUTATIONS, VARIATIONS, AND REI^ATIONSHIPS OF THE OENOTHERAS. 83 
 
 or whether he is engaged in tracing the character, frequency, and causes of 
 mutation a preHminary statistical investigation will aid in the choice of mate- 
 rial best suited to his purpose. It is obvious that the great variabilitv of any 
 elementary species would disfavor it as research material for the study of 
 mutations rather than indicate its desirability, as the wide range of its fluctua- 
 tions would suggest the likelihood of its being a recent mutant rather than a 
 mutating species. A more promising procedure with respect to any variable 
 form would seem to be to use the most nearly related or contiguous species 
 having a low degree of variability in the hope of thus hitting upon the parental 
 type from which the variable form originated. The student of pedigree- 
 cultures frequently receives letters from the kindly disposed, calling his atten- 
 tion to strikingly variable forms which are, on the ground of their variability, 
 conceived to be especially valuable. It is to be hoped that in the future more 
 attention will be given to constant forms having one or more variable near 
 relatives. 
 
 Whether the relative variability of old and new structures could likewise be 
 used by the morphologist as a new criterion for working out lines of descent it 
 would be somewhat reckless to say at present, but an important field for 
 research is indicated. Are forms which are obviously more recent at the same 
 time more variable than the older? Are the structures characteristic of the 
 Orchidaceae, for instance, more variable than those of more generalized mono- 
 cotyledons, and those of the Compositae more variable than those of other 
 dicotyledons? Are monocots on the whole less variable than the dicots? 
 While not wishing to press the application of the principle too far, it may be 
 pointed out that one feature of the investigation here reported seems to bear 
 directly upon the phylogenetic significance of the greater variability of new 
 characters. The characteristic feature of the Oenothera bud is the hypan- 
 thium. It doubtless represents one of the most fundamental mutations in the 
 history of the genus. It accords well with the hypothesis here offered to find 
 that the length of the hypanthium is in every species the most variable part 
 of the bud. The great excess in its variability over that of other parts of the 
 buds is made strikingly apparent by comparing the average variability of the 
 several parts of the buds in all the species studied. The average coefficients of 
 variability are as follows, in order of their magnitude : Length of hypanthium, 
 14.97; length of cone, 9.56; length of ovary, 8.66; thickness of cone, 7.85; 
 thickness of ovary, 6.30 ; thickness of the hypanthium, 5.83. The consistency 
 of these results make it certain that the great variability of the hypanthium 
 originated with the origin of the hypanthium. 
 
 The fact that a new species is more variable than an old one should also 
 give to the phytogeographer a method of investigating the exact status of an 
 endemic species. If it presents a case of initial endemism it should be more 
 variable than the nearly related, wide-ranging, and hence presumably older 
 
84 MUTATIONS, VARIATIONS, AND RELATIONSHIPS OI^ THE OENOTHERAS. 
 
 species. If on the other hand it is a local relict of an old species, which once 
 had a wide distribution, it should present less variability than its wide ranging, 
 presumably newer relative. The only satisfactory evidence heretofore avail- 
 able for determining a point of this kind has been drawn from paleontological 
 sources. 
 
 To the metaphysician who is striving to picture to himself the laws of vital 
 motion, the results of these statistical studies will give an altogether new view 
 of the relation between fluctuation and mutation. Rosa (1899) has published 
 a treatise in which he contends that there is a progressive lessening of varia- 
 bility during the progress of evolution, and he sees as the certain result of this 
 process the ultimate extinction of life. Plate (1904) takes issue with Rosa on 
 this point and maintains that there is no such law. The demonstration that 
 there is greater variability in new species, coupled presumably with a decrease 
 in variability as the species grows older, indicates that both Rosa and Plate 
 are partly right. It is now seen that although the individual species may 
 decrease in variability as it grows older, this decrease is compensated for every 
 time a new species springs into existence. Instead of mutations being the 
 cumulative results of ever-increasing fluctuation, they appear now as an initial 
 process of which fluctuation is in part an after effect. 
 
 A detailed description of but a few of the species of Oenothera have been 
 given. Seeds and material have been brought in from all parts of America 
 east of the Rocky Mountains, and a consideration of this material gives 
 foundation for the conclusion that under the name biennis is included a swarm 
 of elementary species very closely related, but easily distinguishable when 
 grown side by side, having no intergrading forms. 
 
 Even the imperfect knowledge we possess of the species, owing to the 
 scarcity of collections made of them, shows that they overlap in distribution 
 in the most haphazard manner. In the same way, the elementary species 
 comprised within cruciata interlock. In discussing the principles that govern 
 distribution, it is to be admitted that in accordance with their usual practice 
 many systematists would not allow specific rank to the forms in question, so 
 closely are they related. It will be in order, therefore, to take up the next 
 grouping of interest, that of biennis, and the allied species of oakesiana, parvi- 
 flora, 7nuricata, and cruciata, which were accepted as varieties of biennis 30 
 years ago, but which are accorded specific rank at the present time. It is 
 found that biennis blankets nearly the entire range of these forms, and also 
 that the ranges of these forms overlap among themselves. It is to be seen, 
 therefore, that no matter what range of opinion may enter into the discussion 
 as to the rank of the forms of the evening-primroses of the eastern part of 
 North America, the thesis that the most nearly related species do not overlap 
 in distribution is unsupported. 
 
MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 85 
 
 A condition similar to the above has been found in Crataegus by Sargent 
 & Peck, who say (1906) : 
 
 The peculiar tendency of Crataegus to flock together is strikingly illustrated in our terri- 
 tory. It is rare to find any large area occupied by a single species. Where many thorn 
 trees and bushes grow together there are usually many species. A remarkable example of 
 this kind is found in a narrow strip of pasture land bordering the Erie Canal near Menands. 
 Here 10 species are growing in an area of about i acre. The closest condensation of num- 
 erous species that I have seen anywhere is near Albia, where 9 of our native species are 
 growing in a kind of irregular row along the west bank of the Wyantskill Creek. The 
 length of the row is about 100 feet. It is worthy of remark that 3 of these species, Cratae- 
 gus fereiitaria, C. rhombifolia, and C. succulenta, belong to the group Tomentosae. Such 
 close associations of niembers of a single group as this one at Lansingburg are very signif- 
 icant and when more fully understood may possibly throw some light on the interesting 
 problem of the development of species. 
 
 Other students of the thorns have found similar conditions with strains 
 which were found to be distinct and hereditary in pedigreed cultures. Here, 
 again, if a conservative view is upheld and specific rank is not accorded these 
 clearly separable elementary species, a similar association of the closely 
 related group species may be found. A similar condition among the opuntias 
 has been noted by the author. These facts make it conclusive that the 
 assumption that the nearly related or most nearly related species of a group 
 do not occupy the same habitat is untenable. 
 
 Two bud-sports of different types as to origin were found during the 
 summer of 1905 and tested as to their hereditary qualities. 
 
 The vegetative mutant branch of ammophila was seen to give rise to a 
 progeny which came true to the characters of the branch producing the seeds 
 from which they were grown, and the reversion, if it is such, may be consid- 
 ered complete. If, as suggested, ammophila is a hybrid and the sport is a rever- 
 sion to one of the parents, then this case may be taken to agree with that noted 
 by De Vries (1900, p. 86), in v.'hich a hybrid of two forms of Veronica gave off 
 bud-sports, which in purely fertilized seeds came wholly true to one of the 
 parental types. In this case the sport was a reversion to the ancestral qualities 
 which had become latent when the hybridization was effected. 
 
 The second bud-sport was one in which the recessive characters appeared on 
 a branch in the first generation of a cross, and when this was purely fertihzed 
 showed only a strict inheritance of the recessive characters. 
 
 The facts offered by bud-sports have much significance as to the localization 
 of mutations, and as to the nature of the stimuH which set the mutatory pro- 
 cesses in action. Primary buds usually arise laterally from the tip of the 
 growing point and within a millimeter or two from its apex. The saltation, 
 of whatever character it may be, occurred in a cell or cells in the embryonic 
 tissue while the growing tip was moving upward a distance of a millimeter or 
 more. Much depends upon the place where this occurs. If it is possible for a 
 
86 MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 
 
 single cell near the apex of the growing point to undergo such changes during 
 the rapid division that characterizes this region that it gives rise to a mass of 
 cells from which the bud arises, this would favor the probability that the salta- 
 tory changes are due entirely to internal causes and might not be affected at all 
 by circumstances. 
 
 Examples of sports are known in which a section of the shoot involving 
 several branches are of an atypic character. DeVries offers an illustration in 
 his description of a green branch on a colored oak in which the greening not 
 only affected the entire branch, but also a section of the member from which it 
 arose, showing that the mutated protoplast was one which gave rise to more 
 than the atypic branch. In other instances the mutation may involve only a 
 longitudinal section of a branch or of a member or limited growth such as 
 inflorescence, and this may be carried to the extent that even half of a flower 
 may be of an atypic character, and likewise similarly complex fruits may be 
 produced. It is, of course, admissible that in reversionary sports, as most of 
 these sectional variations are, the qualities that appear in the sport are latent 
 in the entire plant and need but a slight stimulus to awaken them, or some 
 force to weaken the dominancy of the main stock. So far as the saltation here 
 is concerned it is one in which any disturbance of the equilibrium would in all 
 probability have but one result. 
 
 In the case of the appearance of characters not inherited in a latent, reces- 
 sive, or active condition, but which constitute a progression or acquisition, the 
 case is different, and if more than one protoplast is concerned in the alterations 
 ensuing preliminary to the organization of a bud-sport, then the suggestion 
 that the alterations were the result of stimulation from factors external to the 
 cell in which the changes ensued would have great force. 
 
 Bud-sports usually occur on the lower part of the main stem, where many 
 of them lie dormant in sleeping buds, and may be awakened by decapitation 
 of the shoot. It is thus to be seen that they belong to a comparatively early 
 stage in the life of the sporophyte. The mutations which give rise to atypical 
 seedlings, on the other hand, seem on theoretical grounds to be due to changes 
 ensuing in the very closing stages of the sporophyte, and previous to the quali- 
 tative or reducing divisions which form the egg, or in the early stages of the 
 pollen mother-cells. Here is encountered a feature not yet recognized in bud- 
 mutations — that of a fairly constant frequency. 
 
 No evidence has yet been secured to show that the pressure or direct action 
 of environmental factors upon the vegetative organs has produced any per- 
 manent inheritable changes in elementary strains of plants. The record of the 
 experiments in which solutions of various kinds were injected into ovaries 
 immediately previous to fertilization seems to point definitely to the con- 
 clusion that the bearers of the hereditary characters may be definitely and 
 directly affected by forces external to the cell, however. The alterations thus 
 
MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 87 
 
 induced consist in the suppression of some qualities of the parental form and 
 the substitution therefor of other characters with which the suppressed 
 features are mutually exclusive. The changes are expressed in anatomical 
 alterations and modifications of attendant functions. 
 
 Atypic derivatives have been secured in this manner from two species, 
 Raimannia odorata and Oenothera biennis, hy the use of more than one reagent, 
 as described in the preceding text. The induced mutants have been tested to 
 the second and third generation, with the result that both are found constant 
 in the sense that they do not intergrade back to the parental form. 
 
 The injection of the solutions into the ovaries results in the total destruction 
 of some eggs by the mechanical action of the needle and of the solution itself, 
 while in a few the stimulative or other action is of such nature that the changes 
 are induced which result in the modification of the hereditary characters and 
 of the adult sporophytes which carry them, while doubtless in most of the 
 cases no action at all resulted. Of course in some of the unsuccessful tests all 
 of the ovules were destroyed and fell off before maturity. While it seems 
 more natural to assume that the injection results in direct action of the fluids 
 on the egg-cells, yet the possibility is by no means excluded that the saturation 
 of the tissues of the ovary with the solution results in the induced effect being 
 due directly to the action of the reagent upon the advancing pollen-tube and 
 its contents. 
 
 The wide variety of substances used as reagents in securing the alterations in 
 question makes it impossible to suggest the nature of the changes induced in 
 the reproductive elements, since the actual effect may be osmotic or chemical. 
 Neither may it be said whether the changes consist in disturbances of the pro- 
 cessions of enzymatic action, or actually modify the structure and composition 
 of the cytoplasm or nuclear constituents. 
 
 Mutants have thus been produced by conditions brought about in an exper- 
 imental way, and it can not be denied that similar action of external forces 
 might occur unaided by man. The radioactivity of spring and rain water, or 
 the effects of corrosive or other gases which are being set free in numberless 
 places on the earth's surface, might produce effects similar to these obtained 
 in the pedigreed cultures. Undue or abnormal secretions from the walls of the 
 ovary, from neighboring tissues, or from' intruding foreign pollen incapable 
 of accomplishing fertilization might exert a similar influence. The stings and 
 lacerations of insects and animals and the action of parasitic fungi are known 
 to be accompanied by the liberation of certain substances, the activity of 
 which results in profound changes in the somatic tissues. It is not unrea- 
 sonable to suppose that the effect of these substances upon the ovary might 
 exert equally important changes in the egg apparatus or developing ovules. 
 In its present state this investigation appears to be of great potential impor- 
 
88 MUTATIONS, VARIATIONS, AND RELATIONSHIPS OF THE OENOTHERAS. 
 
 tancc, since it offers the suggestion that we may become able, more or less at 
 will, to modify existing streams of heredity and cause new organisms to arise. 
 The point not to be lost sight of is that the action of external agents as 
 described and deemed possible is not through the somatic tissues to the egg or 
 germ-plasm, but is exerted directly upon the reproductive elements themselves. 
 To assume that such action is stimulative, as has been done by Tower in dis- 
 cussing the earlier announcement of these results (Tower, 1906), is to adhere to 
 one of a number of allowable suppositions — one, however, which seems most 
 applicable to the results secured by him with beetles, but one to which we are 
 by no means confined in the consideration of the induction of the atypic forms 
 in plants. The determination of the character of this action constitutes one of 
 the most difficult problems in connection with the entire matter. 
 
MUTATIONS, VARIATIONS, AND RE;i.ATIONSHIPS OP THE OENOTHERAS. 89 
 
 RECAPITULATION. 
 
 The principal features of the foregoing paper may be stated in the following 
 brief generalizations : 
 
 (i) The probabihty of the origination of new elementary strains of plants 
 arising by saltations or sporting is one which has received the support of many 
 authors, from inferential conclusions based upon the examination of living 
 material. It has also been found by more than one worker that characters 
 acquired in a saltatory manner are not swamped by intercrossings. If such 
 characters behave as units they may well survive, whether dominant or reces- 
 sive with respect to the corresponding character of the parental stock, the prob- 
 abilities for survival being somewhat greater with recessive characters. 
 
 (2) Frequency of mutation is put forward as a better expression than 
 periodicity of mutation for the proportion of salts occurring in the course of 
 generations. 
 
 (3) An evening- primrose inseparable from 0. lamarckiana occurred around 
 Haarlem, Holland, as early as 1756, and near Liverpool, England, in 1806. 
 It has been noted in this and other districts at various times, and has been 
 under more or less continuous observation since 1892. Apparently estab- 
 lished and growing with it are 0. rubrinervis and 0. lata, two of its mutants. 
 0. hrevistylis, another mutant which is recessive when crossed with O. la- 
 marckiana, has been in existence in competition with the parental form in 
 Holland for twenty years. 
 
 (4) The coefficient of mutability of Oenothera lamarckiana has not been 
 increased in the cultures in America, although a form which has hitherto 
 escaped observation was secured. Seeds of this parental form, from various 
 sources outside of Amsterdam, furnished a smaller proportion of mutants than 
 the material furnished by Professor De Vries. 
 
 (5) 0. ohlonga has been found to constitute as much as 54 per cent of the 
 atypic derivatives of 0. lamarckiana. 0. lata, from material grown in England, 
 was found capable of self-fertilization and gave rise to a progeny containing 
 the same elements as when fertilized by the parental form. 
 
 (6) Fixed hybrids constituting species were secured in combinations of 
 0. lamarckiana and 0. cruciata. 
 
 (7) The greater variability of phylogenetically new characters as compared 
 with older ones, which was supported by statistical evidence brought forward 
 in a previous paper, is confirmed by the evidence presented by further studies. 
 
 (8) The forms which would appear to promise most of importance in a 
 pedigree-culture would be species with a low degree of variability, having 
 nearly related forms with a high degree of variability. 
 
 (9) The hypanthium of Oenothera, which seems to be a recent development, 
 is more variable than any other feature of the bud. 
 
90 MUTATIONS, VARIATIONS, AND RELATIONSHIPS OI? THEj OEJNOTHKRAS. 
 
 (lo) The inference seems justifiable that individual species become less 
 variable as they grow older, and that this decrease is compensated for by the 
 fact that when a new species springs into existence, it has a greater degree of 
 fluctuability than its parent. Such saltations are the starting-points from 
 which series with decreasing fluctuation follow. 
 
 (ii) 0. parviflora, which has been known in Europe since 1759, has been 
 found in its native habitat in America in the same manner that O. grandiflora 
 was traced to its original habitat. This leads to the hope that O. lamarckiana 
 may be found in a wild state. 
 
 (12) The species of evening-primrose of eastern America are distributed 
 in such manner that the most closely related species overlap in distribution. 
 This conclusion is true, no matter upon what taxonomic basis the forms 
 are classified. Similar conditions in Crataegus and Opuntia are cited. 
 
 (13) Two bud-sports have been followed through two generations and 
 found to be constant. The sport in one case embodied the recessive char- 
 acters of a hybrid combination showing alternative inheritance, the recessive 
 qualities being thus activated and extracted in the first generation, coming 
 true thereafter. 
 
 (14) The action of reagents having an osmotic and a chemical effect has 
 resulted in the induction of mutants in the progeny of Raimannia odorata 
 and Oenothera biennis. The mutants thus induced have been tested to the 
 second and third generation and found to come true to their newly assumed 
 characters. 
 
 (15) The induction of mutants by the action of reagents is a conclusive 
 demonstration of the fact that hereditary characters may be altered by 
 external forces acting directly upon the reproductive mechanism. The 
 action of the reagents used experimentally is simulated by many conditions 
 occurring in nature. 
 
MUTATIONS, VARIATIONS, AND RELATIONSHIPS Olf THE OENOTHERAS. Qt 
 
 BIBLIOGRAPHY. 
 
 1789. AiTON, William. Hortus Kewensis: or, A catalogue of the plants cultivated in the 
 Royal Botanic Garden at Kew, 2, 2. London, 1789. 
 
 1902. Bailey, Charles. On the adventitious vegetation of the sandhills of St. Anne's- 
 on-the-Sea, North Lancashire. Mem. and Proc. Manchester Lit. and Phil. 
 Soc, 47: pt. I, 1902. 
 
 1793. B.\RTRAM, William. Travels through North and South Carolina, Georgia, East and 
 West Florida, the Cherokee country, the extensive territories of the Musco- 
 gulges or Creek Confederacy, and the country of the Chactaws. Dublin, 
 1793. (Reprinted from the Philadelphia edition of 1791.) 
 
 1839. Baxter, W. British phanerogamous botany, 4: pi. 257, 1839. 
 
 1859. Darwin, C. Origin of species. 
 
 1900. DE VriES, H. Das Spaltungsgesetz der Bastarde. Ber. d. deut. hot. Gesell., 18: 
 
 83-90, 1900. 
 
 1901. De VriES, H. Die Mutationstheorie. Versuche und Beobachtungen ueber die 
 
 Entstehung von Arten im Pflanzenreich. Erster Band. Die Entstehung der 
 Arten durch Mutation. 648 pp., 8 pis. Leipzig, 1901. 
 
 1902. De VriES, H. Die Mutationstheorie. Versuche und Beobachtungen ueber die 
 
 Entstehung der Arten im Pflanzenreich. Zweiter Band. Die Bastardirung. 
 Erste Lieferung. pp. 1-240. Leipzig, 1902. 
 
 1903. De VriES, H. Die Mutationstheorie. Versuche und Beobachtungen ueber die 
 
 Entstehung der Arten im Pflanzenreich. Zweiter Band. Elenientare Bastard- 
 
 lehre. pp. 241-751, 4 pis. Leipzig, 1903. 
 1905. De VriES, H. Species and varieties: Their origin by mutation. Edited by 
 
 D. T. MacDougal. xviii -f- 847 pp. Chicago and London, 1905. 
 1905. De VriES, H. Ueber die Dauer der Mutationsperiode bei Oenothera lamarckiana. 
 
 Ber. d. deut. bot. Gesell., 23: 3S2-387, 1905. 
 1898. Field, W. L. W. A contribution to the study of individual variation in the wings 
 
 of Lepidoptera. Proc. A. A. A. S., 33: 389-396, 1S9S. 
 
 1904. FOCKE, O. Oenothera ammophila. Abh. Nat. Ver. Bremen, 18: 182-1S6. 1904. 
 1907. Gates, R. R. Pollen development iu hybrids of Oenothera lata X O. lamarckiana, 
 
 and its relations to mutation. Bot. Gaz., 43; 81-1 15,07. 
 
 1902. GrEEX, C. T. The flora of the Liverpool District, p. 58, 1902. 
 1845. Hall, T. B. Flora of Liverpool, 37, 1845. 
 
 1879. HoLLiCK, Arthur. A few notes on the abnormal absence of color in plants. Bull. 
 Torr. Bot. Club, 6: 293-296, 1879. 
 
 1869. KernEr. Die Abhangigkeit der Pflanzengestalt von Klima und Boden. Inns- 
 bruck, 1869. 
 
 1903. M.\cDougal, D. T. Mutations in plants. American Naturalist, 37: 737-770, 
 
 November, 1903. 
 
 1905. MacDougal, D. T. Discontinuous variation and the origin of species. Science, 
 
 21:20-23, 1905- Also Torreya, 5: Jan. 1-5, 1905. 
 1905. MacDougal, Vail, Shull, and Small. Mutants and hybrids of the Oenotheras. 
 Publication 24, Carnegie Institution of Washington, 1905. 
 
 1905. MacDougal, D. T. The evolution of plants by mutation. Lecture before the 
 
 Barnard Botanical Club, New York, December 18, 1905. Chicago, 1905. 
 
 1906. MacDougal, D. T. Discontinuous variation in pedigree-cultures. Pop. Sci. M'thly, 
 
 September, 1906. 
 
92 MUTATIONS, VARIATIONS, AND REI.ATIONSHIPS OF THE OENOTHERAS. 
 
 1882. MeEhan, T. Hybrid oaks. Bull. Torr. Bot. Club, 9: 55, 1882. 
 
 1 760. Miller. Figures of the most useful and uncommon plants described in the Garden- 
 er's Dictionary exhibited on three hundred copper plates accurately engraven 
 after drawings taken from nature, vol. 11, 1760. 
 
 1904. Plate, Iv. Giebt esein Gesetz der progressivenReduktionderVariabilitat? Arch. 
 
 f. Rassen u. Gesells. Biol., 1: 641-655, 1904. 
 1895. PoHL, J. Ueber Variationsweite der Oe«o</;(?ra /amarcHawa. Oesterr. Bot.Zeitschr., 
 
 45: 166-169, 205-212, 1895. 
 1899. Ros.v, D. La riduzione progressiva della variabilita, pp. 135. Torino: Carlo 
 
 Clausen, 1899. (French Abstract, Arch. Ital. de Biol., 33' 314 et seq., 1900. 
 
 German translation by Dr. H. Bosshard.) 
 1903. Rosa, D. Die progressive Reduktion der Variabilitat und ihre Beziehungen zum 
 
 Aussterben und zur Entstehung der Arten, pp. 105. Jena: Gustav Fischer, 
 
 1903- 
 
 1905. Rose, J. N. Studies of Mexican and Central American plants. 4. Contributions 
 
 U. S. Nat. Herb., 8; 330, 1905. 
 
 1906. Sargent, C. S. , and Peck, C. H. Species of Crataegus found within 20 miles of 
 
 Albany. N. Y. State Museum Bulletin No. 105. Report of the State Botanist, 
 
 1905. 1906. 
 1906. ScHAFFNER, J. H. A successful mutant of Verbena without external isolation. 
 
 Ohio Naturalist, 2: 31, 1906. 
 1806. vSmith, J. E. English botany, pi. 1534, 1806. 
 1906. TowEr, W. L. An investigation of the evolution of the chrysomelid beetles of the 
 
 genus Lc pa notarsa. Publication 48, Carnegie Institution of Washington, 1906. 
 
 North Carou. .... .^^^ 
 
r oifr 
 
North Carolina state university Libraries 
 
 Sr^rrVARIATIONSANOBe^TJONSH. 
 
 S02776136 H 
 
 ili%-3. 
 
 ?H^1 
 
 
 ' %y V : 
 
 'K^'^ 
 
 ••a . <, 
 
 n>'^ 
 
 *-i?\1