386 MEMOIRS OF THE NEW YORK BOTANICAL GARDEN 



plant A as a. seed parent was fertile to 4 and sterile to 3; as a 

 pollen parent it was fertile with 3 and sterile with 4. The plant A 7 

 as a seed-parent was fertile with 2 and sterile with 3 of the sister 

 plants, and also sterile with the parent A; as a. pollen-parent it 

 was sterile with i, and fertile with 3 of the sister plants and also 

 fertile with .1. The plant Aiy as a seed-parent was fertile to 4 

 sister plants and sterile to one and with the parent ^ ; as a pollen- 

 parent it was sterile to one sister and to the parent A, which were 

 the only two cases attempted. 



In most cases the data are fully conclusive for positive or 

 negative results, but there were numerous instances of feeble 

 cross-fertility, giving results quite similar to the cases of feeble 

 self-fertility already reported. In some cases, as in the A8 X A 

 combination, where only one head out of thirteen set seed, there 

 may have been an experimental error. The evidence is clear, 

 however, that cross-incompatibility exists among these plants 

 and that among those which are compatible varying degrees of 

 such a condition may exist. 



Table 9 includes data on 16 reciprocal crosses; 7 pairs between 

 parent A with i of its offspring and 9 pairs between sister plants. 

 Nine of these gave similar results; 5 pairs were fertile and 4 pairs 

 were sterile. In 7 pairs, however, the reciprocals were different. 

 These cases are of special interest and are collected in table 10. 



The parent plant A was fertile to the pollen of Ay, but the 

 reciprocal was sterile. \A'hen it appeared from the seed produced 

 that this was the condition, the pollinations oi Ay were continued 

 until 30 heads oi Ay were pollinated with pollen from plant A, but 

 not a single seed was produced. Pollen of plant A was effective 

 on plants C, Ej, and E22 (see table 2) and on plants /Ip and Azy 

 (table 8); also the plant Ay produced seed abundantly to pollen 

 of Aq besides setting abundant seed to open pollination, so it is 

 clear that impotency was not involved and that the compatibility 

 is different for the reciprocals. The results in the case of Ag X 

 /I27 and A8 X Aiy are quite decisive, although a smaller number 

 of heads were tested. In the other four cases the fertility of the 

 one reciprocal was low in that a small proportion of heads set 

 seed; here the fertility was of the grade quite comparable to the 

 frequent cases of feeble self-fertility already reported. It is, 

 however, to be recognized that as already stated such results may 



