1918] JONES—HETEROSIS 331 
The development and structure of the endosperm of angiosperms is so 
much like that of gymnosperms that it seems easier to regard the various 
fusions as merely resulting in a stimulus to growth than to imagine a degener- 
ate embryo assuming this particular development and structure (Joc. cit. 
. 183). 
Considering double fertilization as an adaptation means that 
endosperm hybridization arose as a different process from that of 
nuclear fusion in which nuclei derived from one individual take 
part. Of course the union of like nuclei cannot be considered as a 
means of altering the food supply, so that NEmeEc’s hypothesis has 
no bearing upon this phase of the problem. Neither can the union 
of like nuclei be a means of increasing the amount of food in the way 
that endosperm hybridization does, since heterosis, according to the 
hypothesis recently advanced by the writer (ro), is not due to an 
indefinite physiological stimulus, but merely the result of bringing 
together the maximum number of growth factors showing partial 
dominance.? 
If increased endosperm development is simply a manifestation 
of heterosis and as such can be put on a Mendelian basis, the process 
of endosperm hybridization, in so far as it arose as a means of either 
increasing the amount or altering the kind of food supply, is a 
Phenomenon quite apart from the fusion of like nuclei. Moreover, 
if double fertilization came about as an adaptation, having occurred 
in cross-pollinated plants, it must have persisted as a process of no 
value, both in species which are now almost entirely self-pollinated, 
as well as in those which do not produce an appreciable amount of 
endosperm, as NEMEC points out. 
Whether or not heterosis can be removed entirely from the 
category of results due to indefinite ‘physiological stimulations,’ 
in which category the results of the fusions of like nuclei would still 
be, remains to be seen. Some interesting results obtained from 
wheat crosses have an important bearing on the question. Both 
G. F. Freeman} of the Arizona Experiment Station and K. Sax° 
of the Washington Experiment Station have obtained independently 
f arcoint 
+h ¢ 
2 ae two serious objections tot tl e hy I 
ing for h ah si ck on tanta at kage of hereditary 
factors are bias into Cmaidecation. 
3 Unpublished data. 
