Tricotylotts Half Races. 379 



cases, can he clearly recognized as consisting of two 

 leaves (Fig. 81 D). 



The fact that so rare a teratological phenomenon 

 should occur so regularly in two well-known varieties of 

 the same species, — the one uniform red, and the other 

 striped yellow and red — means perhaps that the char- 

 acter in question has existed for a long time in the Snap- 

 dragon and will be found, after a close investigation, 

 to exist in other cultivated varieties also and possibly 

 even in the wild ancestral form. Of course the fact that 

 I found them in a tricotylous race need not necessarily 

 indicate a causal relation between this character and 

 tricotyly, because, at the beginning of my cultures, I 

 started by selecting the tricotyls and continued the race 

 from their seeds alone. If such a relation did exist the 

 fact that the anomaly occurs both on dicotylous and on 

 tricotylous individuals would be very important, for it 

 would show that it is not the visible tricotyly itself, but 

 some corresponding internal character, which must be 

 regarded as the cause. It is to be hoped that the abund- 

 ance in which the anomaly can now be obtained will 

 render possible a closer examination of this problem. 



§ 4. TRICOTYLOUS HALF RACES. 



Occasional tricotylous seedlings will be found among 

 samples of seed in very many species. All that is neces- 

 sary, therefore, to start a culture is to buy a sufficient 

 quantity of seed and to sow it. The seed will either 

 give no aberrant forms, or very few, or a considerable 

 number. In the first case the possibility of obtaining 

 tricotyls still remains open if a larger quantity of seed 

 is sown. In the second case the variants can be used as 



