STUDIES IN TERATOLOGICAL PHENOMENA 
29 
very surprising, inasmuch as the hybrid family had contained only 
cream and white-flowered plants even to the grandparental generation. 
Table 3 shows the ratio of white to colored plants and their stamen 
character. 
Table 3 
Pistillody not 
Color Pistillody fully expressed Normal Total 
White 71 10 81 
Colored J_ J_ 7 9 
72 II 7 90 
When I found that some of the progeny with sap-colored (magenta, 
etc.) flowers possessed pistilloid stamens, I was more puzzled than 
ever, because I had already found it to be completely recessive in the 
crosses I had made. When the conception of dominance and reces- 
siveness as characteristics, not of the unit "character" or factor alone, 
but of the latter plus the effect produced upon it by its internal 
(genotypical) and external environments, was brought to bear upon 
the problem, the explanation was simple, especially as 90 Fi and 381 Fo 
progeny of a cross between -2-1 A and 321 (N. alata) had given nothing 
but white-flowered plants. During the winter I had been working 
with many colored-flowered F2 segregates of N. forgetiana (314) X N. 
alata (321) and had not been careful enough about cleaning my pollen- 
izing tools before selfing the flowers on the cuttings of the original 
(2-1A) mutant, and, as a result, a few hybrid seeds were produced. 
Pistilloid stamens in the colored-flowered plants were due to dominance, 
complete in one case and partial in the others, of the anomalous condi- 
tion over that of the normal. In the other 7 progeny with colored 
flowers, the expected condition, i. e., the dominance of the normal, 
prevailed. Probably all 18 progeny belonging to the normal and inter- 
mediate classes were hybrid. Further experiments are in progress 
to determine this. The change in dominance is not thought to have 
any special connection with the color factors, but is interpreted in the 
same manner as the anomalous results secured in some of my un- 
published studies on fasciation, viz. : the modifying influence of other 
factors. The 18 plants which were causing confusion had, in the 
majority of cases, a very different and distinct habit from the original 
pistilloid mutant, and this was especially true of the plants with colored 
flowers. The 72 or more pure abnormal (2-1 A) progeny were very 
similar in habit, flower color and other characters, so much so that I 
