As the other portions of the primitive leaf have been so won derfully modified in the course of their development and altered from their original condition, so the freed lateral portions to which we may now apply the term stipules have not retained their primitive proportions in adult leaves nor the identity of all their parts. But as the central basal portion has often almost wholly degenerated, the same thing has happened to the basal parts of the lateral portions. The parallel degeneration of the two por tions has brought the stipules into closer and closer apparent re lation to the stem, so much so as to lead to the enquiry whether they are not accessory leaves and to suggest their origin from the reduction or lack of development of a portion of the leaves as in Selaginella and their subsequent association in close relation to the larger ones, but in all my investigation I have not found the slightest evidence in support of this theory. The degener ation of the stipules may continue until they become vesti gial or finally disappear altogether. This is evidently the ease in those families of plants a few species only of which still possess stipules, as for example the Caprifoliacere.
But opposed to the basal degeneration of stipules, there has very commonly been a longitudinal development to that by which the lamina has been evolved. This has resulted in the adaptation of the stipules to the peculiar requirements of each genus and species. Often in this secondary development they remain membranous, serving the protective function only, and when free are early deciduous. But in numerous cases they have acquired the assimilative function also, developing abundant chlorophyll and sometimes, as in the pea (Pisum sativiunt L.), be coming of equal assimilative importance with the lamina. In Lathyrus Aphaca L., they even replace it almost entirely.
Among all these varying forms we should expect to find closer similarities in those plant groups of nearer relationship as we do in floral structures, and conversely these similarities of foliar de velopment should also point to relationship, clue allowance being made for parallel development in adaptation to similar environ ment and for secondary functional modifications which find mor phological expression. Also in types more recently evolved and more highly differentiated wide divergence from the typical mode of development may be looked for. The Caprifoliaceie, before mentioned, are of such a type, with stipules usually wholly aborted; another is the family of the 11 ubiacere with anomalous stipnlar development in the group of the Stellatre. The oaks also, though of lower organization, are an advancing type and still actively undergoing differentiation as evinced by the close relationship and difficulty of determination of the species of any given group. In this genus all but the upper part of the primitive leaf has disap peared by degeneration even in the earliest stages represented in embr•onic leaf-development, and the well developed stipules are distinct and separate from the very base of a developing shoot. Not until the fifteenth node, in Quercus rubra L. (fig. 39), is there
any appearance of lamina. The apical portion of the protophyll must however be regarded as potentially present between the stipules at their base. It begins its development unusually late in the series and exhibits several stages, of which the twentieth leaf (fig. 40) is illustrative, before reaching adult size. The axial portion of the protophyll being aborted, the petiole, here again a short one, is formed by the contraction of the basal part of the lamina itself. The case of Fagus is very similar, but the lamina appears as early as the eighth node (fig. 41), indicating a less de gree of specialization. In related genera a different course has been followed. The lamina develops still earlier and the stipules of the lowest nodes are united, separating only on the appearance of the first accompanying lamina.
In the family of the uglandaeete the genus Hicoria furnishes a very interesting example. The lower foliar organs are of the primitive type with an unusual development in size in some species. The transition to the adult leaf-form is commonly rather abrupt, but I have observed, in both Ilicoria alba (L.) Britton and IL microcarpa (Nutt.) Britton, the frequent occur rence of intermediate forms, the lateral portions remaining as typical adnate stipules (fig. 42).
I have not seen the typical representation of embryonic leaf de velopment better exemplified than in the case of Baplisia linetoriu (L.) R. Br. where at a glance one is struck with its clear ness. It is also especially fall and accurate as occurring in the development of subterreanean buds. The first five leaves are ex tremely primitive, completely surrounding the node, though only slightly developed on one side. The fifth (lig. 43) shows at its apex a minute apical tooth, the beginning of the lamina which is farther developed in the sixth leaf. I n the seventh (fig. 44) the three leaflets are plainly distinguishable, the petiole has begun its development and the separation of the stipules has made considerable advance. The ninth leaf (fig. 45) is well developed, with the large stipules still showing considerable adnation. But in the tenth (fig. 46) they are wholly free and much reduced, and higher up disappear altogether. We could hardly have a more complete series in illustration of the formation of stipules than this, giving as it does all the stages from an extremely primitive leaf-form to that very highly organized condition where the stipules have entirely dis appeared. By a comparison of the venation in the seventh and ninth leaves, it will appear that the separate condition of the stipules has been attained in the manner already described, partly by the formation of an apical cleft, partly by the degeneration of the central-basal portion bringing the base of the cleft lower down. Meanwhile there has also been a considerable apical development of the stipule itself. But this increase in size is lost again in the tenth leaf and the reduction continues to final abortion. Meli lotus elba Lam. presents very similar though somewhat less primitive conditions.