Bütschli and Fol made the process more generally known. The publication of their researches excited the greatest interest, and a host of observers, amongst whom I may especially name Strasburger, Flemming, Hertwig, Balbiani, E. van Beneden, Johow, Heuser, Pfitzner, J. M. Macfarlane, Carnoy, and Rabl, demonstrated the process in a number of plants and animals, and the literature of the subject is now very extensive. In order to express the appearances presented, and the changes which take place both in the nucleus and in the cell in the process of division, a new nomenclature has been introduced, and we now read of cytaster, monaster, dyaster, equatorial plate and crown, pithode or cask-shaped, spindles, ellipsoids, coils, skeins both compact and loose, pole radiations, spirem, and other terms. From the range of the literature it would be a work of considerable labour and time to make an analysis of the different observations so as to associate with the name of each observer the particular set of facts or opinions which he has made known. Fortunately, this is unnecessary on my part, as admirable resumés of the whole subject have recently been published both by Professor M'Kendrick of Glasgow1 and Professor Waldeyer of Berlin.2

Without entering into a detailed description, it may suffice my present purpose to say that four stages may be recognised in connection with nuclear division.

The first, or spirem stage, exhibits several phases. At its commencement the finer threads, which connect the primary or coarser chromatin fibres of the resting nucleus together, and which give the network-like character, have disappeared along with the knots at their points of intersection and the nucleoli. The primary chromatin fibres, or chromosome as Waldeyer calls them, form a complex coil, the spirem or ball of thread, which divides into loops, about twenty in number, and forms a compact skein. The loops are placed with their apices around a clear space called by Rabl the "polar field," whilst their free ends reach the opposite surface of the nucleus or the " antipole." The nucleus also increases in size cotemporaneously. The loops next become not so tightly coiled, and form the loose skein, though the individual fibres thicken and shorten. A most important 1 Proc. Phil. Soc., vol. xix., Glasgow, 1888.

2 Archiv für Mikros. Anat., Bd. xxxii., 1888.

change then occurs, which was discovered by Flemming, and which consists in a longitudinal splitting of each loop or primary chromatin fibre into two daughter threads. A spindle-shaped figure, first seen by Kowalevsky, next appears in the nucleus; it consists of threads that stain much more feebly than the chromatin fibres.1 The spindle has two poles and an equator, and it finally occupies a position in the deeper part of the nucleus; its equator lies in the plane, through which division of the nucleus is about to occur. The loops of chromatin fibres group themselves in a ring-like manner around the equator of the spindle with their angles inwards, whilst from each pole of the spindle a radiated appearance (cytaster) extends into the protoplasm of the cell. The membrane of the nucleus has now disappeared, so that it is directly invested by the protoplasm of the cell; and it is possible, as Strasburger thinks, that there may be a direct flow of the protoplasm into the nucleus, and that the spindle may be produced by it. At the pole of the spindle, from the point at which the cytaster radiates, E. van Beneden has seen a small, shining, polar body, which Strasburger says is not found in vegetable cells.

The second, or monaster stage. When the chromatin loops. have arranged themselves about the equatorial plane of the spindle with their limbs pointing outwards, and the angle of the loop towards the centre of the spindle, a single star-like figure (monaster, equatorial plate or crown) is produced. The two daughter threads into which each primary chromatin thread had previously split longitudinally, now separate from each other, and, according to Van Beneden and Heuser, pass to opposite poles of the nuclear spindle, where they form loops. These changes are known as the process of metakinesis.

In the third, or dyaster stage, the chromatin loops at each pole of the spindle arrange themselves so that the angles of the loops, though not touching each other, are close together at the pole, and the limbs of the loops are bent towards the equator of the spindle. Two stars are thus produced (dyaster), one at each pole, and each star is formed of one of the daughter threads into which each chromatin fibre of the monaster divides

1 Owing to the feeble staining of the spindle figure and of the nucleoplasm, the substances which compose them have been named Achromatin.

by its longitudinal splitting. Each star is sometimes called a daughter skein; around each daughter skein a membrane appears at this stage, and a daughter nucleus is then formed.

In the fourth, or dispirem stage, the chromatin threads thicken and shorten, and the loops of each star arrange themselves with the angles towards the polar field of the nucleus, and the limbs to the antipole.

The division of the mother cell into two new daughter cells is now completed by the cell protoplasm gradually constricting in the equatorial plane until at last it is cleft in twain, and each daughter nucleus is invested by its own mass of protoplasm. The chromatin threads of the daughter skein then form a network of coarser and finer fibres, a nucleolus appears, and the resting nucleus of the daughter cell is completed. Two daughter cells have thus arisen, cach of which possesses its own independent vitality. Owing to the very remarkable longitudinal splitting of the fibres of the chromosome, and the distribution of the daughter threads from each fibre to the opposite poles of the spindle, it follows that each daughter nucleus contains about one-half of each chromatin fibre, so that whatever be the properties of the chromosome of the mother cell, they are distributed almost equally between the nuclei of the two daughter cells. As regards the cleavage of the protoplasm, there is no evidence that such a rearrangement of its constituent parts takes place as to give to each daughter cell one-half of the protoplasm from each pole of the mother cell. It is probable that each daughter nucleus simply becomes invested by that portion of protoplasm which lies in proximity to it at the time when the constriction of the protoplasm begins. The young daughter cell, seeing that it is composed both in its nucleus and protoplasm of a portion of each of these constituent parts of the mother cell, possesses therefore properties derived from them both.1

Owing to the disappearance of the nuclear membrane at the end

1 Dr J. M. Macfarlane has described as constantly present within the nucleolus of vegetable cells a minute body, which he terms nucleolo-nucleus or endonucleolus. He considers it as well as the nucleolus to become constricted and divided before the nucleus and the cell pass from the resting into the active phase of cell multiplication. See Trans. Bot. Soc. Edin., 1880, vol. xiv., and Trans. Roy. Soc. Edin., 1881-82, vol. xxx.

of the spirem stage of karyokinesis, at least in cells generally (though it is said to persist in the Protozoa during the whole process of karyokinesis), it follows that the nucleoplasma and the cell protoplasm cease for a time to be separated from each other, and an interchange of material may take place between them in opposite directions--both from the protoplasm to the nucleus, as Strasburger contends, and from the nucleus to the protoplasm, as has in addition been urged by M. Carnoy. In every case it should be remembered that the nucleus, being surrounded by protoplasm, can only obtain its nutrition through the intermediation of that substance, and thus there is always a possibility of the protoplasm acting on the nucleus, and in so far modifying it.

Having now sketched the progress of knowledge of the structure of cells and their mode of production, I may, in the next instance, state the present position of the subject. We have seen that the original conception of a cell was a minute, microscopic box, chamber, bladder, or vesicle, with a definite wall, and with more or less fluid contents. This conception was primarily based upon the study of the structure of vegetable tissue; and, as regards that tissue, it holds good to a large extent to the present day. For the cellulose walls of the cells of plants, with their various modifications in thickness, markings, and chemical composition, constitute the most obvious structures to be seen in the microscopic examination of vegetable tissue. Within these chambers is situated the active, moving protoplasm of the cell, and embedded in it is the nucleus; it also contains the sap, crystals, starch granules, or other secondary products. The cell wall is to all appearance produced by a conversion of, or secretion from, the protoplasm. But even in plants a cell wall is not of necessity always present; for, in the development of the daughter cells within a pollen mother cell, there is a stage in which the daughter cell consists only of a nucleated mass of protoplasm, prior to the formation of a cell wall around it by the differentiation of the peripheral part of its protoplasm. Again, the so-called non-cellular plants or Myxomycetes, before they develop their spores,1 consist of masses of naked protoplasm, on the exterior of which, in the course of time, a

1 Lectures on the Physiology of Plants, by Julius von Sachs. Translated by H. Marshall Ward, Oxford, 1887.

membrane or cell wall is differentiated; in the substance of these masses of protoplasm numerous nuclei are situated.1

In animal tissues the fat cell possesses a characteristic vesicular form, with a definite cell wall, but neither in it nor in the vegetable cells does the cell wall exercise any influence on the secretion either of cell contents or of matters that are to be excreted. In animal cells a cell wall is frequently either nonexistent, or doubtful, and when present is a membrane of extreme thinness. Animal cells, therefore, do not have as a rule the chamber-like form or vesicular character of vegetable cells.

The other constituents of the cell, and the only essential constituents, are the nucleus and the material immediately surrounding it in which the nucleus is imbedded. It is of secondary importance whether this material be called protoplasm, or bioplasm, or germinal matter. The term protoplasm, however, is that which has received most acceptance. In adopting this term, it should be employed in a definite sense to express the translucent, viscid, or slimy material, dimly granular under the lower powers, minutely fibrillated under the highest powers of the microscope, which moves by contracting and expanding, and which possesses a highly complex chemical constitution. The term

ought not to embrace either the cell wall of the vegetable or animal cell, or the intercellular substance of the animal tissues. For although these have in all probability been originally derived from the protoplasm, by a chemical and morphological differentiation of its substance, or a secretion from it, they have assumed formal and special characters and have acquired distinct functions. Protoplasm, as above defined, is a living substance endowed with great functional activity. It possesses a power of assimilation, and can extract from the appropriate pabulum the material that is necessary for nutrition, secretion, and growth. Growth takes place not by mere accretion of particles on the surface, but by an interstitial appropriation of new matter within its most minute organised particles. In cases, also, where the media in which the cell lives are suitable, as in the freely moving Amoeba, or the white blood corpuscles, portions of the proto

1 The opinion for long entertained that the simpler algæ and fungi and cryptogams generally are destitute of nuclei has been shown by Schmidt and others to be incorrect.

C