\66 Ganong . — The Comparative Morphology of 
and the less developed the cotyledons, and the lagging of the 
embryos behind the adults is probably true here also. The 
latter principle may be found true where a single species 
of a given genus suddenly deviates from the normal for that 
genus. This lagging behind amply explains why adult and 
embryo do not always correspond. But however it may be in 
details, I think there is no doubt that the ground-form of the 
embryos is imposed by the ground-form of the adults, though 
not so much by the immediately preceding as by the more 
remotely preceding ancestors. 
When we turn to consider the dynamics of the connexion 
between adults and embryos, exactly how it is that the 
former can influence the latter, we face a difficult question. 
At a first glance one might explain it as due to the fact that 
as embryos and adults live in the same places, and hence 
under much the same conditions of dryness ; and since 
condensation or spread of surface is chiefly determined by 
dryness of the habitat, therefore the spread of surface must 
correspond in adults and embryos, i. e. through independent 
adaptation of each to the same external conditions. But only 
a little observation is needed to disprove this : for on the one 
hand the germinated embryos live under very different 
external conditions from the adults, in that they are growing 
only at the time of the yearly rains ; and on the other, Cereus y 
Echinocactus , Opuntia , and Mamillaria embryos grow side by 
side on the same desert and thrive equally well, though of 
such extremely different forms. Direct adaptation therefore 
to the surroundings, except in the case of those with a re- 
acquired mesophytic habit as Phyllocactus , can have but little 
to do with the form of the embryos. Again, one may suppose 
that the shape of the seed determines that of the embryo ; 
but this simply replaces the difficult question at issue by the 
very much more difficult one of how the adult plant is able to 
influence the form of its seeds so that these will mould an 
embryo answering in form to itself. Moreover the relative size 
of cotyledons and hypocotyl is not in the least degree affected 
by the shape of the seed. In some minor particulars however, 
