The plant has an elongated prostrate thallus with a thick and very conspicuous midrib (Figs. 7, A, 8, E). The thallus may reach a length of upwards of 20 cm. with a width of 7 mm. The thallus is usually forked and is attenuated posteriorly, due to the gradual suppression of the lamina. The midrib itself is practically of equal diameter throughout, and projects strongly on the ventral side. As already stated, the wings of the thallus become gradually narrower toward the base of the A. Male plant, x 3. B. Cross-section of young conducting tissue, very highly magnified. than in that species. The best differentiation was secured by Haidenhain's iron-alum-haematoxylin. The blepharoplasts could not be made out, but presumably are present. A sufficient study of the spermatogenesis in P. Levieri was made to show that it is very much like that in the other two species. Fig. 12, S, shows the pair of spermatocytes with the separating membrane and the blepharoplasts. The spermatocytes are somewhat larger than corresponding stages in P. Zollingeri, measuring about 9μ in length. The Archegonium In all species of Pallavicinia the archegonia are in groups surrounded, as already stated, by a double envelope: an outer one, the involucre, much the more conspicuous before the fertilization of an archegonium; and an inner one, which is very small at first, but which after an embryo is formed grows rapidly and forms a conspicuous tubular sheath enclosing the developing sporophyte. In P. Zollingeri, the involucre is cup-shaped, with a lobed margin (Fig. 6, G, H). Within this, and surrounding the base of the archegonial group, is the young perianth, which at this stage does not reach above the level of the venter of the archegonia (Fig. 13). P. radiculosa (Fig. 13, A, B, E) differs from P. Zollingeri mainly in the much greater number of archegonia in the receptacle, and in the more flaring and deeply fringed involucre. P. Levieri is somewhat intermediate in character, both as to the number of archegonia and the form of the involucre (Fig. 13, D). The receptacle is at first level with the surface of the thallus, but as new archegonia develop it becomes raised and forms a more or less prominent elevation, or placenta, to which the archegonia are attached (Fig. 13, E). Of the three species examined, P. radiculosa was the best for the study of the archegonium, as all stages of development were present in the material. P. Levieri also showed most of the stages, but as it differed very little from P. radiculosa an exhaustive study was not made. None of the specimens of P. Zollingeri showed very young archegonia, but to judge from the few immature archegonia that were seen it does not differ essentially from the other species. Pallavicinia radiculosa, while agreeing in the main with other anacrogynous Jungermanniales, in the development of the archegoniurn shows certain differences that may be noted. The youngest archegonia (Fig. 14, A, C) show a stalk composed of two or three cells, and a terminal, approximately hemispherical cell from The growing point of the shoot lies at the bottom of a more or less evident notch, formed by the active growth of the tissue on each side of the apical cell. The latter has segments cut off only from its two lateral faces. Seen in horizontal section (Fig. 8, C) it is a narrow triangle, with A. Cross-section of the thallus apex, showing the apical cell, x, and the apex, b, of a branch, a, a are the young scales subtending antheridia. B. The apex of the shoot, more highly magnified. C. Horizontal section of the shoot apex, showing the apical cell, x, and a very young antheridium. D. Vertical longitudinal section of the apex. E. Archegonial plant, x 3. the sides converging somewhat in front. In vertical longitudinal section it appears also triangular, but very much broader (Fig. 8, D). A crosssection (Fig. 8, A, B) shows that the lateral faces meet above and below at an acute angle, so that the whole cell has somewhat the form of a triangular wedge, with the edges directed respectively dorsally and ventrally. The segments cut off from the lateral faces are large, and ordinarily are formed alternately right and left. Before a dichotomy occurs, however, more than one segment may be cut off successively on one side, and one of these presumably gives rise to the apical cell of the new branch, the branching being in the strict sense of the word not a true dichotomy. In horizontal and vertical longitudinal sections the arrangement of the segments derived from the apical cell is much the same. In crosssection the elongated, almost oval cell appears surrounded by the crescentshaped segments (Fig. 8, A, B). Fig. 8, A shows a cross-section of the thallus apex shortly after the dichotomy is complete, and the growing point of the new branch is established. The lobes a, a', are the beginnings of young antheridial scales, while the group of cells, b, is the growing point of the new branch. The divisions of the apical cell are not in quite the same plane as those of the original apex. Each segment cut off from the apical cell first divides transversely into an adaxial and an abaxial cell. The latter cell divides only by walls perpendicular to the plane of the thallus, and contributes to the wing or lamina of the thallus, which remains permanently but a single cell in thickness. The adaxial cell divides crosswise into a middle and an inner cell. The former forms the outer tissues of the midrib, and also from it arise the antheridia. The innermost cell, by subsequent repeated longitudinal divisions, gives rise to the narrow cells of the conducting strand traversing the midrib. These cells at first have dense contents, which later mostly disappear. A section of the young conducting strand is shown in Fig. 7, B. The entire apex of the shoot, and the younger antheridia, are bathed in mucilage which is secreted by two-celled glandular hairs developed from the outer cells of the young segments; and similar mucilage secreting hairs are also found among the antheridia. Pallavicinia Levieri Schiffner Pallavicinia Levieri, the third species considered here, is much less restricted in its distribution. It is quite common in the neighborhood of Tjibodas in Java, at an elevation of 1400-1500 metres, and material was collected at several points. It has been reported from several other sta tions in Java and Sumatra, and according to Stephani occurs also in the Pacific Islands, Tahiti and Hawaii. It will probably be found in other parts of Polynesia. The plant (Fig. 9) is smaller than P. radiculosa, and there is much more difference in size between the male and female plants. Like P. radiculosa, the thallus is prostrate. It usually occurs on the trunks of trees, among other liverworts and mosses, and does not form masses of large size. The plants are very delicate in texture, and the male plants are noticeably smaller than the females or the sterile plants. They |