SEEDLING STRUCTURE IN THE LEGUMINOSJE. 89 
Part С. 
GENERAL DISCUSSION. 
INTRODUCTORY. 
Diverse views are expressed by different writers on the value of seedling 
anatomy as an index to phylogeny. While some authors are disposed to the 
opinion that “16 is not justifiable to use the seedling structure as an indication 
of phylogenetic connections” (de Fraine, 1910, p. 175), others, for example 
Sargant (1903, p. 5), base far-reaching conclusions on the resemblances 
found in certain cases between the seedling anatomy of species from distinct 
families. It is necessary to treat such evidence with caution, for vascular 
anatomy stands in a very intimate relation with physiological needs ; and 
except in eases of obviously vestigial characters there is usually an explanation 
alternative to that of relationship aud heredity. In all discussions of kinship 
as indicated by anatomy it is essential to distinguish the cenogenetic from 
the palingenetic ; and not until the possibility of a direct explanation in 
terms of physiology has been dealt with can we safely ascribe resemblances 
to the conservative influence of heredity. 
In the morphology of seedlings we find a singular uniformity which 
contrasts sharply with the multiformity of habit of the mature plant. In the 
anatomy, too, there is usually no indication in the seedling of the peculiarities 
of specific structure which will be exhibited in later life. This is a striking 
phenomenon in the Leguminose, and with certain exceptions, е. g. the 
Caetaceze (Ganong, 1898; de Fraine, 1910), appears to be a general rule. 
Two problems are involved, one with regard to morphology, the other to 
anatomy. The morphological problem is how far the uniformity of the 
seedling form reflects the uniformity of environment and physiological needs 
in embryology and early life, and how far it is reminiscent of an ancestral 
type. The answer to this question has almost invariably been to aseribe 
almost everything to the former alternative. Not so in the case of the 
similar anatomical problem. It is confidently expected in some quarters that 
ontogeny in vaseular structure will prove to repeat phylogeny, and that in 
seedlings anatomical resemblances are likely to indicate relationship between 
species on divergent branches of a family tree. 
It must be admitted that such a prospect is highly improbable, and the 
search for phylogenetic relationships through the medium of seedling anatomy 
a forlorn hope. We are confronted by the relative unreliability of vascular 
anatomy as a key to phylogeny; and by the admitted fact that the seedling 
morphology of the Phanerogams owes its uniformity to similar cenogenetic 
influences. The problem is not the same for the Phanerogams as for such 
