with in considerable numbers and a distinct nucleus may be detected. Very soon however by the aid of reagents, especially the aniline color methyline-blue', a delicate thin transparent hyaline membrane or wall is found to clothe and to have been formed all over its surface by the protoplasm, which has in some cases where the preparation has been treated with alcohol slightly contracted away from this wall. The membrane is however exceedingly difficult to detect at this stage. This change takes place simultaneously in all the special-mother-cells. The newly formed cell, consisting of a very thin and delicate cellulose wall, closely applied to the internal side of the pale yellow cuticularized wall of the special-mother-cell by which it is surrounded, but from which it may be made to contract away by means of alcohol, enclosing protoplasm loaded with vacuoles and rendered dark with minute granules, and a nucleus, is the equivalent of the pollen-grain of other plants, and will in future to indicate this feature be designated by the same title. The mode of formation of the pollen in Asclepias is very different to that which is the characteristic and prevalent type in the majority of Dicotyledons or Monocotyledons; and, so far as our present knowledge extends, exhibits in its entire details a perfectly unique, isolated and peculiar case of development. The earlier stages are only to be found paralleled in the single instance of Zostera, which affords either the most primitive or most aberrant type of pollen-formation known. The later stages find no precise parallel in the entire range of the vegetable kingdom. This is the more remarkable since another member of the Asclepiadeæ, viz. Periploca graeca, exhibits according to Reichenbach, a type of pollenformation exactly comparable to that of the Orchid genera Neottia and Epipactis. Observations on the mode of development of the pollen in Asclepias are fraught with extreme difficulty, and its history can only be revealed by careful study of extremely thin transverse and longitudinal sections. In many of the pollen-grains, especially when the flower was fully mature, I was able by careful observation and by having recourse to osmic acid of one per cent. strength and to staining reagents such e.g. as Haematoxylon, Grenicher's carmine, and 1 I owe the suggestion that I should make use of this staining reagent to my friend Mr W. Gardiner, who has employed it largely in his researches on "The continuity of the protoplasm in the motile organs of leaves." 2 Dr 8. H. Vines suggests that probably in Asclepias and likewise in Zostera the phase of the special-mother-cells as it occurs in other plants is omitted, and hence we get the marked departure from the normal types. On this view what I have termed the special-mother-cells are really the last series of mother-cells produced by repeated division of the single primary one. some of the aniline colours, viz. Gentian-violet, Saffranin and Methyl-green (to the latter of which a few drops of solution of Acetic Acid, one per cent. strength, had been previously added), to detect not a single nucleus only, but two nuclei one of which was invariably larger than the other. The smaller nucleus was often found lying close to the cellwall, and in these cases I believe that, surrounded by a small quantity of protoplasm, it is cut off from the rest of the grain by a cellulose-wall although I was not always able to shew this satisfactorily. This discovery is especially of interest in connection with the recent researches of Strasburger1 and Elfving' since further confirmation of their observations has thereby been obtained in the pollen-grains of plants which they did not investigate, and in which the very presence of nuclei of any kind whatever had not been previously detected. I consider the smaller nucleus of the Asclepiad pollen-grain to be the representative of what Elfving terms the "vegetative nucleus" and others have dignified as the "passive nucleus," which nucleus is genetically the last remnant of the male prothallium of a Vascular Čryptogam type such as Equisetum; while the larger nucleus, equivalent to the "active nucleus", is genetically the last remnant of the antheridium of such a type. In shape the pollen-grains are always nearly spherical, though usually slightly angular, so as to be really irregularly polyhedral; their membrane, is as previously stated, single, very thin at first, ultimately becoming thicker, smooth hyaline and transparent, and formed of unchanged cellulose. There is at this stage no appearance whatever of the tubes which are afterwards produced. Strasburger, in his most recently published work, mentions the fact that he has observed the presence of only a single coat in the pollen grains of the following plants-Gaura biennis L., Clarkia elegans Dougl., Senecio vulgaris L., Coboa scandens Cav., Allium L. Naias major"?, and Orchids, and the same phenomenon was described by Fritzsche, and has long been known to occur in 1 "Ueber Befruchtung und Zelltheilung." Jenaische Zeitschrift für Naturwissenschaft, Bd. xi, Neue Folge, Bd. iv. 1877, Heft iv. page 450. 2 Jenaische Zeitschrift für Naturwissenschaft, 1879, part 1., and Quarterly Journal of Microscopical Science, N. S. Vol. xx. 1880, pp. 19 to 35. I have never seen it present the irregular internal thickenings which are frequent in the pollen-grains of some plants, e.g. Cucurbita, and which are there used subsequently in the formation of the pollen-tube. • Ueber den Bau und das Wachsthum der Zellhäute, Jena, 1882. * Ueber den Pollen.-Mêmoire présenté à l'Académie impériale des Sciences de St Pétersbourg, iii. 1837. • Hofmeister (Neue Beiträge, 1859, part ii.) describes the existence of "a very thin but distinct extine" in the pollen-grains of this species, but in his figure, pl. i. fig. 11, he represents this extine as extending with the intine along the course of the pollen-tube produced from the grain! It is therefore highly probable that Strasburger's observation is more accurate. Zostera L., while Asclepias Cornuti Dcne. must now be added to this interesting list of exceptions to what is the otherwise universal rule in Phanerogamous plants. I may here allude to the circumstance that after extremely careful observations many times repeated I have at length in several cases satisfactorily traced the passage for some distance of the larger of the two nuclei in the pollen-grain into the long cellulose tube which it puts forth when the pollinium ruptures in consequence of its being placed in immediate contact with the stigma, although I have been unable to follow its changes further. The larger nucleus of the pollen-grain does not then in this genus become broken up and diffused through the protoplasmic contents of the grain immediately before the production of the pollentube, as Strasburger has shewn that it does in many other flowering plants, but simply passes in its entire concrete form into the pollen-tube. What becomes of the smaller nucleus I cannot definitely state, but I have reason to believe that it is left in the grain, or perhaps dissolved, since I have never been able to trace its passage into the pollen-tube, although I have carefully watched for it. The ultimate changes and fate which the tapetal membrane undergoes appear to be as follows: : The cells composing it which lie on the outer side of the anther divide, each by means of a vertical tangential wall, parallel to the original tangential walls of the cell, so that the membrane becomes two cells broad on this side. Those tangential cell-walls, which are farthest from the pollinium, in that row of limiting cells which is next the cavity of the loculus, together with the adjacent portions of the radial walls of these cells become broken down and disintegrated. On the other hand the tangential walls which are nearest the pollinium, together with the internal1 portions of the said radial walls, persist for some time forming a continuous membrane surrounded by a layer of small cells. These latter are on the outer side of the anther, segments from those cells derived from the parenchyma that completed the tapetum proper on this aspect, and on the inner side of the anther are the row of cells formed immediately external to the tapetum proper at the same time that it was differentiated and which have persisted. Such is the condition immediately prior to the opening of the two loculi to expose the pollinia, the method of which is intimately connected with the presence of other contrivances in the flower for ensuring pollination by the agency of insects. This change occurs by almost the whole of the parenchymatous tissue forming the substance of the anther, together with the remains of the 1 With relation only to the loculus. 66 tapetal membrane and the upper portion of the inner epidermis which is never cuticularized becoming broken down; an external wall of several layers, and also in the lower two-thirds of the loculus an internal, alone persisting, while in the upper third the pollinium is freely exposed on that side which is directed internally, a large oval longitudinal but slightly obliquely directed aperture in the anther-wall being formed opposite each. This disintegration appears to take place first at the apex of the loculus and gradually to proceed downwards. The apices of the pollinia are thus exposed first, and each pollinium immediately on its exposure becomes attached to the free end of one of the appendages" of a corpusculum which directly impinges upon it, and the end of which is still in a semifluid highly viscid condition. The attachment takes place just below the apex of the pollinium, which is pear-shaped, the narrow end of the pear being directed upwards and slightly obliquely away from the median line of the anther, since the loculus in its superior portion comes actually to abut by one end upon the extreme external wall of the anther itself. It has the form of a minute adhesive surface formed by the flattening of the viscid end of the "appendage" against the external coat of the pollinium. The two parts though externally united are however never confounded. The pollinia being thus firmly held from above, the rest of the parenchymatous tissue disintegrates, and the pollinia are left hanging freely suspended in the two open pouches of the anther, and in no way adherent to any portion of its substance; the pair are separated only by the median vertical dissepiment of the anther which persists connecting the inner and outer sides of the anther between the two cavities. Jacquin', who examined the anthers only in their adult condition when they had already dehisced in this introrse fashion, naturally regarded the anthers in which the pollinia lay freely immersed as "antheriferous sacs" and the pollinia themselves as the true anthers; in this he was followed by Kölreuter and many others, and when Schreber' in 1789 insisted that these sacs of Jacquin were really anthers, he was instantly denounced by an indignant host of authorities, although his opinion has been since most amply confirmed and borne out. The comparatively late period to which the tapetal membrane persists in Asclepias is a noteworthy point: in other Dicotyledons it usually breaks down in consequence of the growth of the pollen-grains after the absorption of the walls of their special-mother-cells; while in the group of the Monocotyledons it becomes either diffluent or ab 1 Selectae Stirpes Americanae, 1763, p. 82. sorbed at an carly period and the mother-cells themselves in consequence float freely about in the loculus quite separate from one another. Asclepias therefore appears to present at first sight in the period of the resolution of its tapetum a closer analogy to the Monocotyledons than to the group of which it is a member; since the pollinium, which consists among other parts of the persistent though altered walls of the mother-cells, comes ultimately to lie in the cavity formed by its resolution. Inasmuch however as the period of its resolution is coincident with that of the dehiscence of the anther-loculus I believe that it more closely approaches the type of its own group than that of the Monocotyledons, though it differs from both so far as we know of them at present. It is further an important feature that there exists in the anthers of Asclepias no provision which shall determine their dehiscence, such as takes place in other plants by the reticulate thickening of the walls in a layer of cells immediately internal to the epidermis and extending round a portion of the area of the loculus, i.e. the so-called "endothecium" of Purkinge'. Schleiden's statement (loc. cit.) that the portion which disappears is dry and elastic, and is cast off as a valve is obviously inaccurate on this point. The nearest instance which I have been able to find to the type which they exhibit is a case described by Hofmeister in which the anther-lobes open at the apex by a pore which results simply from the destruction of a small portion of tissue at this spot, but probably other instances also will not be wanting when we have more extended knowledge on this point, for as yet very little indeed has been done in determining exactly the various modes in which the dehiscence of the anther takes place in different plants. (4) On some micro-organisms and their relations to disease. By G. F. DOWDESWELL, M.A. 1 De Cellulis Antherarum fibrosis. |