points of interest which are still a matter of controversy.

According to many earlier observers, and especially clearly enunciated by the pioneers of the cell theory, Schleiden and Schwann, it was generally assumed that cells arose by a sort of crystallization out of an unorganized ground substance the "cytoblastema" of Schleiden. This idea, however, was soon overthrown by the more complete observations of von Mohl, Unger, Naegeli, Remak, Kölliker, and others, and the foundation was laid for the important generalization of Virchow," omnis cellula e cellula," which has since become one of the most fundamental principles of biology. Every cell is derived from a pre-existing cell by a process of division, and this process has gone on unceasingly from the time when life first began down to the present moment. All life comes from pre-existing life, and, whatever may have in some past time occurred, the spontaneous production of living substance from a non-living condition does not now exist.

Cell multiplication.

One of the earliest results of the study of cell multiplication was the discovery that division of the nucleus precedes the division of the cell body. Furthermore, a careful examination of the different phases of the process offers the strongest proof that the most important feature of this division, an end to which all the other processes are subsidiary, is the exact halving of a certain nuclear substance, the chromatin, between the two daughter cells which result from the division. To gain a clear conception of this process of inKaryokinesis. direct cell division, or "karyokinesis," let us consider the changes which take place in typical cell multiplication. Two parallel series of changes occur nearly simultaneously, the one affecting the nucleus,

the other the cytoplasm. In the so-called "resting" nucleus-i. e., the nucleus not in active division-the chromatin, as we have seen, exists usually in the form of scattered granules arranged along the linin network, and does not colour readily with nuclear stains (Fig. 5, A). As division approaches these chromatin granules become aggregated together in certain definite areas, forming usually a convoluted thread or skein, which now readily takes up the nuclear stains which may be used. In some nuclei this skein is in the form of a single long filament, in others the chromatin is divided up from the first into a series of segments, a condition which soon follows in the case of a single filament (Fig. 5, B). By transverse fission the latter breaks up into a

The chromosomes.

series of segments, the "chromosomes," the number of which is constant for each species of animal or plant. Thus in the common mouse there are twenty-four, in the onion sixteen, in the sea urchin eighteen, and in certain sharks thirty-six. The number may be quite small, as, for example, in Ascaris, a cylindrical parasitic worm inhabiting the alimentary canal of the horse. Here the number is either two or four, depending upon the variety examined. In other forms the number may be so large as to render counting exceedingly difficult or impossible. In all cases, however, one fact is to be especially noted -viz., the number is always an even one, a striking fact which finds its explanation in the phenomena of fertilization to be discussed later on.

While the chromatin is collecting into the form of the chromosomes the nuclear membrane has disappeared. The chromosomes soon reach their maximum staining capacity, and appear usually as a collection of rods or bands of deeply staining substance lying free in the cytoplasm (Fig. 5, C).

Division of the centrosome.

While this is taking place in the nucleus, another series of changes has been gone through with by the centrosome and the cytoplasm immediately surrounding it. We have already indicated the presence of the centrosome as a minute spherical structure lying at one side of the nucleus. This body assumes an ellipsoidal form, constricts transversely into a dumbbell-shaped figure, and divides into two daughter centrosomes, which at first lie side by side but soon move apart (Fig. 5, A). Around each of them is gradually developed a stellate figure composed of a countless number of delicate fibrils radiating out in all directions from the centrosome as a centre. This " aster" or "astrosphere" is at first small in extent, but grows in size progressively as the two centres move apart, apparently being derived from a rearrangement and modification of the thread-like network of the cytoplasm under the influence of the centrosomes (Fig. 5, B and C).

Between these two asters, which lie a short distance apart and at one side of the nucleus, a spindle-shaped system of delicate fibrils may often be

The spindle.

made out, stretching from the centre of one aster to that of the other. This fusiform figure is termed the "central spindle" (Fig. 5, D). The two asters, together with the central spindle, form what is termed the "amphiaster" or the "achromatic" portion of the karyokinetic figure. The two series of changes in nucleus and cytoplasm, which have thus far gone on apparently independently of each other, now become closely interrelated in that, as the nuclear membrane disappears, a system of fibrils grows out from each astrosphere which attach themselves to the individual chromosomes (Fig. 5, D). These "mantle fibres" insert themselves along the chromosomes in such a way that

FIG. 5.-Cell division in the Salamander. (After Drüner.) A, resting nucleus stage, centrosome below divided. B, skein stage, the chromatin visible as a convoluted band; the daughter centrosomes have separated. C, the nuclear membrane has disappeared, a few of the chromosomes lying free in the cytoplasm. D, central spindle complete, the chromosomes, already splitting, are being drawn to the spindle. E, metaphase. F, anaphase. The chromosomes are drawn to the poles.

each segment receives a series of fibrils from each pole of the amphiaster, the two series being attached along opposite sides of the chromosomes. Under the influence of these fibres, probably by direct pulling, the chromosomes, now bent into V- or U-shaped loops, tend to place themselves in a circle around the centre of the spindle, transversely to its long axis, and form the "equatorial plate" (Fig. 5, E).

Splitting of the chromosomes.

The changes thus far constitute the "prophases" of the division. The "metaphases" following these consist primarily in the longitudinal splitting of each chromosome and the moving apart of the halves. This longitudinal splitting of the chromosome into two equivalent parts forms the most important act of the whole cell division, and is of the greatest theoretical significance. By it the chromatin substance of the original nucleus is equally distributed between the two daughter nuclei, so that each receives a half of each original chromosome. The elaborate mechanism and consequent expenditure of energy involved in this careful longitudinal division of each chromosome, rather than a simple mass division, such as might be brought about by far less complicated means, indicates clearly that the distribution of the definite organization of the chromatin to the daughter cells is of primary importance, a conclusion which is further strengthened by much evidence too extended to be entered upon here.

In the "anaphases" and "telophases," which include the closing stages of division, the daughter chromosomes migrate along the fibres of the central spindle toward its poles, perhaps through the direct contraction of the mantle fibres under the influence of the centrosome, though this and many other points regarding the forces at work must be left for future investigation