No. 520] 



EVOLUTION IN VIOLA 



2:1 r> 



In May, 1906, Mr. Stone sent me from Tinicum, Pa., a 

 living violet plant that was quite fertile and appeared a 

 good species. I could not make it out a hybrid, though 

 perhaps predisposed at that time to place an anomalous 

 form in this category. For further light I visited the sta- 

 tion with Mr. Stone the following September; but the 

 most careful search failed to reveal another specimen. 

 It soon afterward occurred to me that it might be an off- 

 spring of V. affinis X sagittata, a hybrid which I had 

 found at the same place the year previous, and had trans- 

 ferred to the garden. The two plants, mother and sup- 

 posed daughter, appeared much unlike, the former being 

 quite infertile, and in most respects an excellent interme- 

 diate between the putative parents that grew near by. But 

 careful examination showed that, though no one character 

 of the supposed daughter was intermediate as in the 

 mother, yet all were to be found in one or the other of the 

 supposed grandparents. The leaves had the breadth and 

 the rounded basal lobes of V. affinis, but the length and the 

 attenuate apex of V. sagittata; the capsules were pubes- 

 cent as in V. affinis, but green and large as in V. sagit- 

 tata; furthermore, the peduncles were strictly erect as in 

 V. sagittata, not ascending as in V. affinis; and also the 

 seeds were brown as in V. sagittata, not buff as in V. 

 affinis. 



The next move was to raise offspring of the plant, to 

 discover if some one of these characters was not impure 

 —in other words, dominant and holding latent the op- 

 posed character. The 23 plants subsequently raised 

 showed all the characters of the plant under investigation 

 to be pure excepting two, the pubescence of the capsules 

 and the dark-brown color of the seeds ; for glabrous cap- 

 sules and buff seeds appeared in some of the offspring. 

 The ratio of the four Mendelian forms in the 23 plants was 

 12: 4: 4: 3, or 9: 3: 3: 2}, an unexpectedly close approx- 

 imation to the theoretical ratio 9:3:3:1. 



From eleven of these plants another generation of 204 

 plants was raised the past season. Among these, in addi- 



