The condition found in Potamogeton is almost exactly repeated in Polygonella articulata (L.) Meisn. (fig. 31). The ochrea is cylindrical, surrounding the stein. central-basal portion is long and narrow, bearing at its apex the terete lamina which is deciduous before flowering. The lateral portions form the prin cipal part of the sheath, are parallel veined with a few anastomos ing bundles and are prolonged above the central-basal portion, growing in along the ridge between it and the lamina. This middle portion shows its origin by a deep median sinus and receives its bundles typically as tangential branches from those entering the lamina. We do not have then in Polygonella a typical ochrea as it occurs-in Rumex and Polygonum, where, because of the small de velopment of the central-basal portion, the sheathing petiole is very short or almost wholly wanting. The lamina, being of much greater importance than in Polygonella, receives all the bundles of the leaf-trace. They are more or less abruptly deflected into the true petiole, generally developed in these genera, according to the degree of degeneration of the central-basal portion. The lateral portions receive their supporting bundles as branches of the lateral ones of the leaf-trace. In Polygonum sagittatum L. (lig. 32), the marginal tissues do not extend across the petiole and we have a stipule opposite the leaf. In _Rumex crispus L. (lig. 33) 811(1 Polygonum Virginianum L. (fig. 34), the ochrea is complete and the axillary parts receive the typical tangential bundles.
The ochrea of palms is doubtless of the same character, though I have not had opportunity to examine its anatomical structure. In those species which I have examined morphologically, the case is that of the ochrea associated with a remarkable development of the sheathing petiole. There is no true petiole and the ligule may be seen even a little above the base of the blade on the upper surface of the midrib. From this point the lateral portions may be traced clown the margins of the sheath, though dried up and very much torn and broken by the more rapid development of the central tissues, till they unite with those parts which in their development have formed the " ochrea." The degeneration of the sheathing petiole with the probable concomitant formation of a true petiole would give the same conditions as in Polygonum with its typical ochrea.
The ochreate stipule of Platanus differs little morphologically from the typical ochrea, except in the absence of development of the central-basal portion and the possession of a horizontal limb, but there is no fibro-vascular support for the ligular part and this usually splits, leaving apparently a single stipule opposite the leaf.
The case of the tendrils of Smilax is one which has occasioned much discussion, but the embryological together with the anatom ical characters make it sufficiently clear that in Smilax the ten drils are true stipules found in connection with the sheathing petiole. If a young shoot of Smilax rotundifolia L. be examined, the first leaf (fig. 35) is seen to be of the typical primitive form. In the second (fig. 36), the apical portion has developed into a blade of considerable size and there is a well-marked sheathing petiole. In the third (fig. 37), the true petiole has begun to de
velop, the central-basal portion is degenerating and at the same time the lateral portions have begun to separate, forming rudi mentary tendrils which in the adult leaves come to considerable length by secondary development in adaptation to their new and unusual function of support. In cross section the bundles of the tendrils are seen to arise as branches of those of the petiole, so that anatomically, as well as embryologically, they answer to true stipules.
Pastinaca satire L. (`fig. 38) furnishes a good example of the sheathing petiole among the Umbellifene. The lateral portions are broad and furnished with several vascular bundles parallel with those of the central basal portion. The lateral portions remain of considerable breadth to the top where they are distinctly rounded off, and their bundles, with the exception of two or three of the exterior ones, curve around and unite with those entering the petiole. This free condition of the exterior lateral with the anastomosing network between them shows a consider able degree of approach to the true stipular condition.
In the third case true stipules are developed. They are formed by a very early separation of the lateral portions from the main body of the primitive leaf, a separation which can be very clearly traced progressively in the embryological history of leaf develop ment. The function of the lateral portions in their primitive connection with the main body of the foliar organ is, in common with the other portions, protective, and while the apical portion, having had placed upon it the special function of assimilation, goes on in its development together with the accessory axial por tion in adaptation to this purpose, the lateral portions usually serve their ancient function only, sharing it with the central-basal portion when this has not disappeared by degeneration. The central-basal portion also supports the main body of the leaf, a function from which the lateral portions have been freed by sepa ration. .
It is in consequence of this separation that all the main vascu lar bundles of the leaf-trace in the third type of leaf-development are deflected toward the central one that they may pass up through the petiole into the lamina and give the required support to these important parts. The support of the lateral portions is left to comparatively small lateral branches from the two exterior bun dles of the trace, evidently developed expressly for the purpose. This we may conclude, since vascular tissue is the most modern of plant tissues and introduced because of the necessity of sup port in the evolutionary development of the primitive ground tissues. It would, therefore, follow and not precede the evolution of leaf-forms, being introduced where needed and disappearing again when degeneration or other support of particular parts ren ders its presence unnecessary. This will appear in some of our examples. In the first and second types of leaf-development the lateral portions may retain in greater or less degree their inde pendent venation.