562 Journal of Agricultural Research voi. xii, no. w 



tributed in the dichotomies of gametogenesis in the Fj generation; if (6) it follows 

 from (4) and (5) that Fj zygotes may be formed which are less perfect in their gamete 

 forming mechanism than those of the Fj generation; and if (7) the heterotypic division 

 of gametogenesis does not necessarily form two cells alike in their viability. 



In the strawberry, in which no irregular distribution of chromatin 

 occurs, certain of these assumptions — namely i, 2, 3, and 7 — would apply 

 in explaining partial male sterility in many partially sterile varieties, 

 but assumptions 4 and 5 and assumption 6, which is dependent upon 

 them, can hardly be considered applicable; assumption 4 because there 

 is no cytological evidence that there has been any disturbance of reduc- 

 tion division; 5, because there is apparently equal distribution of the 

 mother cell cytoplasm to each member of the tetrad, and 6, because 

 there is no cytological evidence which would lead one to believe that the 

 cytoplasm of the zygote had anything to do with the perfection of its 

 gamete forming mechanism. In the strawberry sterility seems to be 

 due to the inability of certain chromosome combinations to use the food 

 material in which they are embedded in the growth and development of 

 the liberated spore to a ripe pollen grain. 



If we can accept the hypothesis of pollen abortion being due in hybrids 

 to certain chromosome combinations affecting the normal metabolism 

 of the microspore in its development, the question at once presents 

 itself: Is pollen abortion the result of the presence of one particular 

 chromosome or of certain combinations of two or more, or do all of the 

 chromosomes play some part in it ? This question can not be answered 

 from the facts so far obtained in the strawberry because of the heterozy- 

 gous condition of the material which has been used; but there is other 

 published evidence which throws some light on this question. 



Belling (4) has made a very careful study of partial sterility of hybrids 

 between four species of the "bean" Stizolobium. He found that the F^ 

 plants, of those crosses in which the velvet bean {S. deeringianum) was 

 used as one parent continually, aborted one-half of the pollen grains and 

 one-half of the egg sacs. Of the second generation plants one-half were 

 completely fertile and one-half partially sterile, as in Fj. The progeny 

 of fertile F2 plants continued to be fertile, while the progeny of the 

 partially sterile plants were one-half fertile and one-half partially sterile. 

 Belling explained these results on the basis of the presence in the velvet 

 bean of the factor K, which was not present in the other three varieties. 

 These three, however, contained the factor L not present in the velvet 

 bean. The presence of either K or L he assumed to be necessary for the 

 normal development of either egg sacs or pollen grains, the presence of 

 both K and L causing abortion. We may extend this working hypoth- 

 esis slightly and put it on a chromosome basis, in which case we must 

 consider that Belling's factors K and k form one allelomorphic pair, 

 being situated in a certain loculus of a specific chromosome in the velvet 

 and the other three varieties of beans, respectively, and that the factors 



