
866 THE AMERICAN NATURALIST. |. (Vor. XXXVI. 
individual, stages may be found in localized parts, which are 
similar to stages found in the young and the equivalent of 
which are to be sought in the adults of ancestral groups.” 
This principle may be well demonstrated in a greater or less 
completeness in each of the following examples. 
Viola pedata L. (Bird’s-Foot Violet). 
Under this species Gray’s Manual gives, “leaves all 3—5- 
divided, or at the earliest only parted, the lateral divisions 
2—3-parted, all linear or narrowly spatulate, sometimes 2-3- 
toothed or cut at the apex," thus recognizing the fact of the 
differences in the leaves but with no significance of the various 
forms or connection between them. 
On April 18, 1902, a study was made of the leaves of this 
species in adult plants which were then well started. Among 
the larger and stronger plants it was found that the earliest 
leaves went back in the accompanying series usually to Pl. I, 
Figs. I0 or 11. After further search plants were found with 
the leaves on the outside of the rosette, similar to Figs. 9, 8, 
or7. Noticing that the smaller plants seemed to be simpler or 
more primitive, attention was turned to them, and some small 
ones were found with the outside leaves going back to Fig. 6, 
and the smallest to Figs. 5 or 4. These were along a woods 
road in a somewhat dry, sandy place. In the middle of this 
road, where there was very little vegetation, were found plants 
evidently very young and weak, with the result that some were 
obtained with the simple leaves shown in Figs. 3, 2, and I. 
That these were not seedlings was shown by the rootstock. 
They were probably weak plants of the preceding year. 
In nearly all the stages shown many specimens were obtained 
of each stage, and in all cases enough to warrant the form as 
not abnormal Where the outer leaves were most primitive 
the more complete were the stages before the typical leaves 
were produced in the center. Seedlings of this plant were not 
obtained, but the earliest of the stages given are very similar 
. to corresponding ones of Viola tricolor L. in both young and 
adult. Later stages would naturally be specific. The seedling 

