942 



SCIENCE 



[N. S. Vol. XXV. No. 650 



2. I would speak, secondly, of diversity in 

 hybrid seedlings as resi)ects pubescence. When 

 one parent is pubescent and the other glab- 

 rous, the hybrids of the first generation, such 

 as we usually find in the field, are as a rule 

 intermediate ia this character, being some- 

 what pubescent. In certain large colonies, 

 where the plants appear to have been long es- 

 tablished, and to have produced offspring, we 

 find notable reversions. I have visited sev- 

 eral stations where V. fimhriatula and V. 

 sagittata grew in abundance, and where many 

 plants were to be seen having the leaf -outline 

 of V. fimbriatula with the glabrous character 

 of V. sagittata; and conversely, many having 

 the leaf-outline of V. sagittata with the hairi- 

 ness of V. fimhriatula. Among cultivated 

 seedlings of hybrids this reversion was seen 

 in V. fimhriatula X septemloha, referred to 

 above. The plants from Long Island and 

 from Connecticut are all more or less pubes- 

 cent; but among the five seedlings that I 

 raised from the Long Island plant, two that 

 in leaf-outline revert to Y. fimhriatula ara 

 quite as glabrous as V. septemloha, retaining 

 only the fine ciliation that appears on the 

 margin of the leaves in both these species. 



3. As respects diversity in size, I have a 

 notable instance in the seedlings of V. papili- 

 onacea X villosa, two species that are respect- 

 ively the largest and the smallest plants of 

 the group. The mother plant was collected 

 near Philadelphia by Mr. Witmer Stone. The 

 five seedlings that I raised from it grew side 

 by side under the same external conditions; 

 but three of the plants had leaves twice as 

 broad as the leaves of the other two. 



4. One more particular in which I have 

 found hybrid seedlings of Viola to differ from 

 one another is in the color of the seeds and of 

 the cleistogamous capsules. In about one half 

 the species of this group these capsules are 

 commonly blotched or dotted with purple. 

 The hybrid of any of these with a green- 

 fruited species bears ordinarily a capsule of 

 an intermediate color. But in the offspring 

 of the hybrid the capsule is frequently seen to 

 have the pure green of the one grandparent, 

 or the normal purple of the other. 



But this color reversion is more strikingly 



exhibited in the behavior of the seeds of some 

 of these hybrid offspring. Our species of 

 ' blue stemless violets ' vary much in the color 

 of the seeds; and these colors are quite con- 

 stant in the same species, as seen in specimena 

 growing a thousand miles apart. For ex- 

 ample, V. cucullata and V. papilionucea have 

 dark brown, almost black, seeds; V. fimhri- 

 atula has nut-brown seeds; Y. afflnis, Y. vil- 

 losa' and Y. septemloha seeds of a light straw- 

 color. Now, when two species with different 

 colored seeds are crossed> the color of the seeds 

 of the first-cross is usually a mean between 

 the colors of the parents. In a hybrid from 

 Ivy Hill Cemetery, Philadelphia, black-seeded 

 Y. papilionacea is crossed with pale-seeded 

 Y. villosa, and produces brown seeds. But in 

 seven plants raised from these brown seeds, 

 four had the dark-colored seeds of Y. papili- 

 onacea and three the light-colored seeds of 

 Y. villosa. In five seedlings of one plant of 

 Y. affinis X cucullata, two bore seeds quite a« 

 pale as those of Y. affinis; the remaining 

 three, however, bore not the black seeds of the 

 other grandparent, but the brovm seeds of the 

 mother-hybrid. This divergence in seed-color 

 also appears in the seedlings of the two hy- 

 brids used to illustrate diversity in leaf-pat- 

 tern. In ten seedlings of Y. cucullata X sep- 

 temloha, four ripened dark-colored seeds; six, 

 light-colored seeds. In five seedlings of Y. 

 fimhriatula y^septemloha, three bore the brown 

 seeds of fimhriatula and two the straw-colored 

 seeds of Y. septemloha. 



I am not yet able to state definitely what 

 proportion of violet hybrids produce hetero- 

 geneous offspring. In some cases the number 

 of plants raised was too small to admit of a 

 satisfactory conclusion regarding this tend- 

 ency. But in several instances there were 

 strong indications that the hybrid was stable, 

 and produced offspring quite like itself. The 

 most noteworthy instance was that of Y. 

 affinis X septentrionalis, of which I raised 

 twenty-four seedlings in 1904, and the past 

 season from the seeds of these, many plants 

 of a third generation. The meager pubes- 



^The species here referred to is the V. villosa 

 of recent authors, which is probably not V. villosa 

 Walter. 



