94 ROBERT TRACY JACKSON ON 



later leaves becomes comparatively sharp, as shown in different leaves of the specimens 

 figured. The tliird nepionic leaf of Fig. 11 has slight emarginate lobes, the beginning of 

 the lobes wliich are characteristic of the species. Gradually during later growth the typical 

 form of leaf of the species is built up. In young trees, up to six or ten feet in height, the 

 leaves, as observed in many cultivated specimens, are large and \ngorous and four-lol)ed 

 (PL 16, fig. 1), no six-lobed leaves having l)een observed in such individuals. 



In the adult tree the tyjjical leaves are four- or six-lobed (PL 16, fig. 12) . Six-lobed 

 leaves are figured in Professor Sargent's Silva as the characteristic leaves, and although 

 they are in some incUviduals uncommon or absent, they ax'e usually abundant in adult trees, 

 as far as my observations have gone. It should also be said that four-lobed leaves are 

 abundant on all typical trees. When four- and si.\-lol)ed leaves occur on the same branch, 

 the six-lobed leaves are on the proximal end, and four-lol)ed leaves on the distal portion, 

 showing this definite relative position in the cases observed. 



In this species the leaves are very variable, but in all variations the distal truncation 

 is never departed from, this distal portion of the leaf (as noted in many other cases) always 

 retaining the character wliich is e\'inced in the young nepionic leaves. Variations are 

 especially noticeable in late or feeble growths, especialW at the ends of branchlets, but 

 may occur as isolated individual leaves on any part of the branch. In PL 16, fig. 13, is 

 represented a characteristic variation of the tip of a branch from the same tree as Fig. 

 12. Leaf 1 has two lateral sinuses moderately developed, a little more so than leaf 3 of 

 Fig. 11, but similar to the characteristic form seen in somewhat older seedlings. Leaf 2 

 has very slight lateral sinuses, and leaf 3 is a close approach to leaf 1 of the seedling 

 (Fig. 11). In PL 16, fig. 4, is represented a small reversionary leaf from the same tree, 

 which is similar to leaf 4 of the seedling (Fig. 11). On the same tree a branch had been 

 mutilated by browsing cattle, and a weak late growth (July) was thrown out. Two of these 

 leaves are represented in PL 16, fig. 7. One leaf has slight marginal sinuses, the other ' 

 is entire, both closely resembling leaves of the young seedling (Fig. 11). All of these 

 aberrant- leaves are seen to be distinctly reversionary in character, closely resembling the 

 condition seen in an earl}' stage of the seedling. Accorchng to the views expressed, these 

 may be considered as leaves which have undergone an incomplete development and are 

 reversionary througli a failure in their individual ontogeny to develop full specific 

 characters. 



In adult Lirlodendron tullpifera leaves are occasionally^ found which have eight lobes 

 (PL 16, fig. 3), there being two small marginal lobes in addition to the six lobes character- 

 istic of the species. Holm ('90) has also described eight-lobed leaves, and my figure is 

 taken from that author. These leaves have developed more than the full specific char- 

 acters and would tend to show what may be expected as the next step in the evolution of 



