served measures 13 cm. in length, carries 11 "nodal disks," and has its axis cut nearly radially throughout its entire length. As here displayed the central tissue is found to be made up of cells which are longitudinally elongated, with end walls often but not universally rectangular, figs. 2, Pl. I., and 3, Pl. II. The cell-walls are for the most part thin except at and in the neighbourhood of the "nodal disks” where they are thicker, and the cells themselves somewhat shorter. In the last quotation from Williamson it will be noted that the central tissue is said to vary very much in different specimens. Some variation is shown by the preparations under consideration, but it only affects the wall thickenings, the size of the elements, the completeness of the tissue, and other minor matters. In no case is the center found to be occupied by vascular tissue when cut exactly in the radial and longitudinal direction. PRIMARY VASCULAR BUNDLES. Returning to the transverse sections, there is another feature to be noticed which is a still more significant one. This consists of (usually) three empty spaces in the tissue which abuts on the periphery of the pith. They vary in size in different specimens, and are sometimes irregular at the margin as if the tissue round them had been torn. These cavities when well preserved at once recal the intercellular passages met with in the primary vascular bundles of Equisetum and Arthropitys, and there can be no doubt that they are strictly homologous with them. In other words, they are the carinal canals of the primary vascular bundles of the axis, and their presence affords positive proof that the bundles are of the Calamitean type. Singularly enough, one or more of these canals are vaguely indicated in some of the figures of transverse sections previously published, but no one has recognised them as characteristic structures. Williamson certainly mentions them in his tenth Memoir,† where he describes them as "small vacant spaces which.. doubtless transmitted vascular bundles to the sporangiophores," but he failed to identify them as the intercellular canals of imperfect bundles, and so missed their significance and homologies. * Ante, p. † Op. cit., Pt. x., 1880, p. 503. In fig. 1, Pl. I., these canals are well and distinctly shown at a, but they appear to have been enlarged by the tearing of the peripheral tissue, as they are smaller in all the other sections. At the margin we observe small elements, e, somewhat different from the rest, projecting into the open space, which forcibly remind us of what is found in the carinal canals of Equisetum. This and other similar sections are of themselves almost conclusive of the nature of these open spaces, but the longitudinal one shown in figs. 2, Pl. I., and 3, Pl. II., is absolutely demonstrative. Here we have one of the canals cut longitudinally, e, and at various points along its course we find, adhering to its sides, fragments of annular, and spiral vessels almost exactly as they are seen in longitudinal sections of the stems of living Equiseta, fig. 3, v. Pl. II. There are remnants too of the tissue which originally interrupted the continuity of the canals at the "nodal disks," one of which is shown at d. I am convinced then that we have here structures which are strictly homologous with the carinal canals of Equisetum and the extinct Arthropitys. Taking the view of Count Solms, that in these two genera the canals represent the initial tracheal strands of vascular bundles, we reach the conclusion already stated, that in these specimens of Calamostachys Binneyana the axis not only had a parenchymatous pith, but that the primary vascular bundles were of the Calamitean type. An examination of numerous other sections, while confirming these conclusions, reveals one or two points of difference which it may be well to mention. In some the pith has a triangular rather than a circular section, and in these cases the carinal canals are situated at the angles. The angles however are usually truncated, and are not always of the same breadth tangentially. Sometimes indeed they give the impression that they are about to bifurcate, the carinal canals dividing, so to speak, at the same time. Obviously we have here an increase in the number of the primary bundles, and in this way the number of canals may be increased to six, a number actually found in two sections belonging to Mr. Lomax, of Radcliffe, which he permitted me to examine. In another case, five transverse sections have been cut from the same strobilus, four of which have three carinal canals, while the fifth has four. As the sections cannot have been far apart in the uncut strobilus, this would seem to show that the number of carinal canals, i.e., of primary vascular bundles, may vary in different parts of the axis. SECONDARY XYLEM. Taking the carinal canals to be the representatives of the primary vascular bundles, fig. 1 shows that in some examples of Calamostachys Binneyana no secondary thickening is met with, though its subsequent appearance is not precluded. In others, however, secondary xylem, composed of scalariform tracheids is met. with, and there as in the stem of Arthropitys, the development begins at the carinal canals. Thus three wedge-shaped masses are usually produced, whose elements spread out in radiating rows and in a fan-like manner from the carinal canals, while a little later intermediate xylem is developed in the intervening areas. In this way a narrow but complete zone of secondary xylem is formed. The three bands of intermediate xylem are usually convex towards the pith, which thus takes the triangular shape with concave sides so often noticed by previous writers. All this however has been so well described and illustrated by Williamson in his fifth and tentht Memoirs that it need not further detain us. * THE CORTICAL TISSUES. The tissue which immediately surrounds the primary vascular bundles or the secondary xylem when such is present, is usually absent from the petrified strobili, and only one section has been published in which it is preserved. This we owe to Williamson, who in his account of it tells us that "the vascular axis is closely surrounded by a dense cellular layer," which "passes into a more open and delicate cellular tissue in which there are large lacunæ probably due to partial desiccation." Fig. 1, Pl. I., represents a section in which the cortical tissues apparently still retain their original form unchanged, In it the inner parenchyma (i.p.) is seen to be composed of cells of relatively large size, with delicate. Op. cit., Pt. v., 1871, p. 72, fig. 38. + Op. cit., Pt. x., 1880, p. 504, fig. 16. Op. cit., Pt. x., 1880, p. 503. thin walls, a circumstance which readily accounts for the frequency with which it disappears in the process of fossilisation. On the outside of the inner parenchyma we have the hypoderma, h, which is thickwalled, and whose elements are somewhat uniform in size, and in some cases have apparently possessed copious contents. In different specimens and probably in different parts of the same specimen the character of the hypoderma varies a little, especially in the wall thickening and the density of the cell contents. As has been shown by previous authors, the hypodermal elements are elongated longitudinally and are prosenchymatous. In none of the sections is there shown a distinct epidermis, but in one or two instances hairs are met with between the sporangia which seem to have been connected with an epidermal layer. THE NODAL DISKS AND BRACTS. According to Carruthers the disks are composed of large roundish cells and the bracts of smaller and more elongated elements, while a slender vascular bundle enters each bract from the axis. Williamsont describes the disk as consisting of "two kinds of coarse thick-walled cellular tissue," broad lines of cells elongated in the plane of the disk leading from the axis to the base of each bract, while the intermediate triangular areas are occupied by a coarse parenchyma. In a later Memoirt he tells us that the prosenchymatous cells are found only on the upper surface of the disk, and that "the bundle of small spiral vessels" runs "along the center of the lines" they form. To these statements we are in a position to make one or two important additions. The disks of one of the best preserved specimens have a thickness of 0559 mm., and the vascular strands run about midway between the upper and lower surfaces. But near the periphery a layer of cells is differentiated beneath the vascular strands which neither Carruthers, Williamson, nor any other writer appears to have noticed. It consists of large elements of variable size, whose walls are thin and whose lumina are occupied by dense masses of some black substance which is probably carbonaceous. *Loc. cit., p. 350. + Op. cit, Pt. v., 1871, p. 60. Op. cit., Pt. x., 1810, p. 503. The histology of the bracts has scarcely received the attention. it seems to deserve, and is certainly more complex than has hitherto been made apparent. A reference to fig. 4, Pl. II., which represents a transverse section, will show that the tissues are well differentiated. At e we have the epidermis, which can often be followed round the whole section at b is seen the sclerenchyma or prosenchyma which is in continuity with that of the nodal disk; and at the delicate strand of vascular tissue. Attention should be specially directed however to a layer of much larger elements which occupy the lower half of the bract, and form one of the most conspicuous of its features. Some of these elements are unusually large, while their walls are comparatively thin, and their contents dense and carbonaceous. They appear to be united without intercellular spaces, and at some distance from the wall a thin pellicle is usually present in each cell, which may represent the primordial utricle. The carbonaceous mass is within this pellicle, and in most cases there is an interval between them, while in others they are in contact with one another. Occasionally both are in contact with the cell wall. This layer of tissue I have come to regard as an important one. It is a constant character of the bracts when perfect, and has been found of great practical value in the recognition of fragments of bracts when detached and in other ways. It appears, however, to have been easily destroyed, as many bracts are met with in which it is wholly or partially lost. It scarcely needs to be said that it is a continuation of the similar layer mentioned above as making its appearance at the periphery of the nodal disk. Longitudinal actions of the bracts are hardly ever met with in a form suitable for illustration. Still by carefully studying the numerous fragments that occur in most preparations of the strobilus it is easy to make out that the elements beneath the upper epidermis are really sclerenchymatous, and that the layer with the black contents is made up of somewhat elongated cells with rectangular or oblique ends. THE SPORANGIOPHORES. The sporangiophores consist of a pedicel and a peltate head or scutellum, and, as shown by Carruthers and Williamson, their his |