152 Rhodora [Jury 
forms. Indeed all organic forms are either directly or remotely ances- 
tral, the parents being ancestors one generation removed. Of several 
species having a common origin some are likely to keep original traits 
longer than others, and do, as is well known in some cases, maintain 
the aboriginal organization almost entire and unchanged for immense 
periods of time. The required ancestral type may therefore often be 
found in existent species; what is occasional — youthful, senile, or of 
sporadic occurrence — in one form, being the habitual and character- 
istic condition in some relative. 
This test of the reversionary nature of a particular youthful character, 
like the odd-shaped leaf of the Barberry seedling, must, as will be seen 
in the case of Sagittaria, be used with caution. In this case, how- 
ever, the resemblance between the youthful leaf of B. vulgaris and 
the mature leaf of B. repens (Fig. 4) as regards both the length of the 
petiole and the shape of the terminal leaflet, seems to be well explained 
by supposing that the leaf of B. repens is very nearly like that of the 
progenitor of both B. repens and B. vulgaris; and that in youth our 
common species takes back to that progenitor, restoring the old form 
to the terminal leaflet, and the original rachis, but not the lateral 
leaflets. 
Berberis vulgaris is a European and Asian form. B. repens isa 
western plant. B. Zhundergit is Japanese. Going to South America 
we find an interesting species for our present purpose in Z. Agapa- 
tensis. ‘The specimens examined were from Bolivia. Fig. 8 reproduces 
the outline of some of the mature leaves. In all but the somewhat 
lessened length of the petiole they agree with the seedling leaf of B. 
vulgaris. Most of the leaves of this species, however, show a little 
tendency to acuteness at the base. 
Berberis TÀhunbergii is particularly curious as to the marginal 
teeth. The first set of leaves after the cotyledons, about four or five, 
are entire, and devoid of spinous processes. Following these are an 
equal number rather coarsely toothed, the teeth bristly-pointed. The 
remaining leaves of the first year become more and more like the 
characteristic leaf, which again is smooth-margined, being without teeth 
and bristles. The smooth leaf, next to the cotyledons, does not seem 
to be an adaptation to any circumstance of the seedling period, but 
rather a reversion to a type possibly older than that represented in the 
infancy of B. vulgaris. The succeeding rough leaves, again, have no 
apparent adaptive relation to the seedling period, and probably revive 
