368 MEMOIRS OF THE NEW YORK HOTANICAL GARDEN 



known thai the (luestions of tVrtiHty" wore in\()l\ccl. The data 

 pertain onh' to the heads from which seed was collected. The 

 successful results of reciprocal crosses show that tiie self-sterility 

 observed in these jilants was not due to impotence of either pollen 

 or egg cells. 



It should be noted that differences in the number of seeds pro- 

 duced per head, such as are shown in table 2 and in all later tables, 

 are not due to the number of flowers in the heads. In 19 14 counts 

 were made of the flowers in 153 heads borne on plant A ; the number 

 \aried from 15 to 22 with the mean at iS. Data on 283 heads of 

 plant B ga\"e a frequency distribution ranging from 13 to 20 

 with the mean at 15.8. A total of 395 heads of plant C gave a 

 range from 13 to 22 flowers per head with the mean at 18.9, As 

 shown in table 2, the seeds collected per head varied on plant A 

 from o to 18; on plant B from o to 17, and on plant C from o to 20. 

 In these crosses the depollination method (01i\er '10) was used 

 and some of the failures may \er\- well be due to injuries or to 

 failure to make the proper pollination of all pistils. It is evident, 

 however, that there is a rather marked degree of fertility in these 

 crosses. The number of complete failures per head is low and 

 there are about half the heads that give a large number of seeds 

 and some heads produced close to the maximum number possible. 



Self-stcri/ity and self-fertility among phuits of the Fi generation. — 

 In the simimer of 1913, a crop of plants was grown from seed 

 obtained by crossing the plants A, C, Ej, and E22. As these 

 plants came into bloom they were tested for self-sterility and the 

 investigations were continued during the summer of 1914 in the 

 attempt to test tjuite adeciuateh- e\ery plant that li\ed through 

 a flowering period. The results are given in table 3. 



Of the 172 plants reported in table 3 there were 157 (over 

 90 per cent) that were completely self-sterile. Fifteen plants 

 set some seed to controlled self-pollination: some of these as 

 (£j X A) no. 4 and {E22 X A) no. 10 were decidedly self-fertile, 

 setting seed in the greater nuiuber of heads; eight plants were 

 self-fertile in less degree, in that seeds were obtained in relati\ely 

 few heads (for example, see data for {A X E22) no. 2 and no. 4) 

 and in some cases as in {A X E22) no. 4 the number of seeds per 

 head was also low; there were five plants each of which produced 

 but one head with seed. 



