large primary segment (fig. 3, L.), and smaller ones derived from the small primary segment (v. fig. 3, S.). If, then, we rely upon the only character at present possible, that is to say, upon the size of the segment, we are bound to conclude that not only does the process of segmentation proceed with great irregularity, but that also there is no evidence of the descendants of the larger primary segment ever forming a cap and subsequently enveloping the descendants of the smaller primary segment.

I believe, rather, that descendant segments of the larger primary segment become intermingled quite irregularly with the descendants of the smaller primary segment, for this undoubtedly affords a better explanation of the appearance of embryos like that shown in fig. 12 or fig. 15.

Fig. 11 is an outline drawing of a specimen taken from a rabbit which was killed at the completion of the 39th hour. In this animal seven of the eight embryos found were in the eight-segment stage. It is extremely difficult to measure the segments with sufficient accuracy to be of any service, and although I measured them I shall not give the results. In this case, as shown by the fig. 11, there was very little dif ference in size, and none, so far as I could judge, in texture. There was one curious feature which is worth mentioning, but I do not attach any importance to it. The larger polar body was visible between the embryo and the zona radiata. The smaller of the two was inside, or rather mingled with the segments of the embryo. It is quite possible that this frequently occurs, for the polar bodies seem very often to disappear entirely. It would be of interest to determine whether, when this is the case, the polar bodies, ever under the altered conditions, acquire a renewed activity and give rise to segments whose descendants become part of the embryo.

I have as yet no evidence as to which sphere commences the next series of cell division.

At about the 47th hour the embryo has the typical morula form, and is made up of a number of segments (sixteen to twenty), which very frequently present great diversity in size.

Nor is there any regularity, as far as I have observed, as to the location of the large and small spheres. Figs. 12, 15, and 17 show this very well.

Fig. 12 was drawn with camera while the specimen was still fresh in a drop of aqueous humour of the rabbit.

After the drawing had been made the specimen was placed in Perenyi's fluid and subsequently freed from the zona pellucida with fine needles. The segments were then separated one from another. In all there were seventeen segments, and of very different sizes.

Fig. 13 shows three of the segments thus separated. Except in size I could detect no difference.

Another specimen from the same Fallopian tube was placed in per cent. solution of silver nitrate for two minutes, and after having been washed in water and exposed to the sun for a few hours was embedded in paraffin and cut.

Fig. 20 is from a section through about the centre of this specimen. The nuclei of the individual segments are not at all distinct, excepting where they have been cut almost through their centres, as no other stain except silver nitrate has been used.

(Of all methods of fixing the early segmenting stages, I believe none answer so well, as far as concerns the preservation of the correct shape of the spheres, as a weak solution of silver nitrate allowed to act for not more than two minutes. Next to silver nitrate I believe osmic acid, 2 per cent., is best.)

Fig. 15, which is a specimen from a rabbit killed sixty-six hours and a half after coition, exhibits in a very remarkable way the great difference in size which may sometimes occur between the several segments.

From the forty-fifth to the seventieth hour the segmentation proceeds slowly, and, I am inclined to think, sometimes very irregularly, as shown by the last-mentioned specimen (fig. 15).

In sections of these stages I do not notice anything particularly remarkable, except that I have completely failed to find any constant character whereby the inner cells can be distinguished from the outer. Fig. 21 is of a section through the

centre of a rabbit embryo aged seventy-six hours and a half. This was preserved in Perenyi's fluid, stained with borax carmine, cut and mounted in series. The one I have drawn is the fourth of eight which pass through the actual embryo itself. The embryo at this stage is composed of a number of cells or segments, as far as I can see all similar in character, though varying a good deal in size.

The cells in the centre are no doubt pressed closely together in the living state, the several clefts, with the possible exception of one, being artificial. This just-mentioned exception, the more regular and continuous slit marked C. BL., is in all probability the first commencement of the slit which ultimately enlarges into the cavity of the blastodermic vesicle.

In only one specimen have I found certain cells to take the stain better than others. I have drawn two figures (18 and 19) from the series of sections in which this occurs, in order to show how such an appearance as that described and figured by van Beneden (pl. iv of his paper) may arise, and at the same time to show how the interpretation put on it by him cannot be held to be sound.

In this specimen the embryo has contracted very much, and is lying quite free from the zona pellucida and albuminous layer.

In fig. 18 the majority of the cells at the surface are seen to be slightly darker than the mass inside, with the exception of two at the point x. This is undoubtedly like van Beneden's figure of an optical section (pl. iv, fig. 1). But if we look at another section, fig. 19, we find here that again the cells of the surface layer have mostly stained darker, with the exception of one at x'.

Hence we are obliged to believe that in this specimen there existed at least two of van Beneden's blastopores, for the light-coloured cells which show at the surface in fig. 18 are at almost the opposite pole to that at which the lighter-coloured cells of fig. 19 show at the surface.

In sections of other specimens of about this age (seventytwo hours) fixed with osmic acid 2 per cent., I have failed also

to find differences of any value between the inner and outer cells. The same may be said of sections of specimens treated with silver nitrate.

Fig. 16 is a drawing through the centre of a specimen from a rabbit seventy-two hours after coition, fixed with nitrate of silver per cent., exposed to the light and stained with picro

carmine.

Summary up to the Seventy-second Hour.

The original description of the segmentation stages by Bischoff I believe to be in the main correct. I cannot find any evidence to support van Beneden's view of the origin of the inner mass of cells from a smaller, more darkly staining primary segment only; or for the origin of the outer layer of cells from a larger, more lightly staining primary segment only; or, again, for the growth of the descendants of one of the two primary segments round the descendants of the other primary segment.

It must be borne in mind that the very fact of van Beneden's description of the later stages of development (the origin of the mesoblast and hypoblast from the inner mass, and that of the epiblast entirely from the outer layer) having been shown to be wrong by several authors (Kölliker,1 Heape, Balfour, &c.), made it almost certain that van Beneden's description of the segmentation stages was incorrect, or, at any rate, that the interpretation he put upon the supposed facts could no longer be held to be sound.

My account is briefly as follows:

1. The ovum about the twenty-fourth hour after coition divides into two segments, one of which is usually larger than the other, there being much variation in this respect.

2. Each of these segments again divides about the twentysixth hour after coition, each dividing very nearly at the same time. These four segments now resulting may vary in size.

3. The third series of divisions takes place about the twenty1 A. Kölliker, Festschrift zur Feier des 300 Jährigen bestehens der JuliusMaximilians-Universität zu Würzburg,' 1882.

eighth hour after coition. There is now less unanimity of action in point of time of the division of the several cells, so that it is fairly common to find embryos with five or seven segments. Here, again, four cells may be distinctly smaller than the remaining four, or all may be almost exactly equal in size. There is no difference to be detected in the character of the contents of the spheres.

4. The segments continue to divide with less and less regularity, so that the descendants of one primary segment become mingled with those of the other, there being much difference in size between the several segments.

5. The result of this continued activity is the formation of a solid "morula " of cells of varying size but similar character, and it is impossible to apply the term epiblast or the term hypoblast to any part of the embryo as yet.

The ovum when it leaves the ovary is enclosed within a sheath or protective investment, the zona radiata. As the ovum proceeds down the oviduct this protective investment is further strengthened by the deposition on its outer surface of a thick coat of some albuminous substance which is secreted by certain cells of the epithelial lining of the oviduct, and which forms a tough strong membrane which has an important influence on the future mode of development.

CHAPTER II.

THE FORMATION OF THE BLASTODERMIC VESICLE.

The Fourth Day (73rd to 96th Hours).

The most noticeable feature of this day's events is the commencement of a cavity within the morula, which cavity enlarges enormously.

What the object of this cavity is we can pretty safely infer, as also we can pretty safely conclude that the causes which bring it about originated in the organism in connection with the diminishing size of the ovum of the distant ancestral animal; but how this cavity is actually produced in the