154 Sargant . — The Reconstruction of a Race of 
function, and no doubt this is very often true. The few careful observations 
on the method of food-supply to the growing embryo indicate, however, that 
there is far more variation in the process than was commonly supposed. 
Two examples illustrate this. 
Professor D. S. Johnson (48) gives good reason to think that in Pepe- 
romia the food is supplied to the growing embryo, as to the young seedlings, 
through the cotyledons. These organs have merely, however, to absorb the 
food in a soluble form. The work of drawing on the reserves laid up in 
the perisperm, and preparing them for use, is performed by the scanty 
endosperm. 
Similar results are recorded by Miss Gibbs (28) for the Alsinoideae, 
with one important variation. Here, as in Peperomia , the endosperm 
prepares the food which it draws from the perisperm for use by the embryo. 
But in the Alsinoideae — in Stellaria media , for example — the suspensor is 
for a long time the absorbent and digestive organ. When the digestive 
function is assumed by the endosperm, that of absorption falls to the root-tip. 
Much more work is needed on this question before we can hope to 
understand the morphology of the embryo. If the cotyledons commonly 
absorb food from the endosperm and pass it on, their early appearance in 
the history of the embryo is sufficiently explained. 
3 . Finally, the development of the embryo is modified by conditions 
other than those which act on it directly. The embryo of the ripe seed is 
not always mature. But whenever the embryo becomes mature it shows 
signs of adaptation to its past and present environment, to the environment 
of its immediate future, and to that of a future rather more remote. We 
have already touched on its adaptations to the conditions surrounding it 
during the period of growth which has just ended. The next stage will be 
the exit of the embryo from the seed, and certain features in the mature 
embryo can be referred to the difficulty of getting free quickly and without 
damage from the tissues which envelop it in the seed. The peg of gourd 
seeds is a well-known example (21, p. 22 and Fig. 5 ), and the action of the 
wedging roots in the Maize (74, p. 11 6 and Figs. 6 - 8 ). Finally, the mem- 
bers of the young seedling are present in the mature embryo, and their 
structure even in the undeveloped state has a certain reference to the 
conditions which will be encountered by the seedling. The coleoptile of 
Grasses, for example, is clearly adapted to pushing up through the soil. 
The conditions, then, to which the embryo must conform are sufficiently 
complicated to account for wide departures in structure from the ancestral 
type. The question before us is whether those conditions would be 
likely to produce temporary suppression of the axis in a Monocotyledonous 
embryo, leading to the formation of an apparently terminal cotyledon. 
This question cannot as yet be answered directly, for lack of evidence. 
It is hopeless to attempt to appreciate the effect of conditions so imper- 
