Now these changes (figs. 531,532.) are very extraordinary and peculiar, and in all animals depend essentially upon these facts. The blood globules first become at once smaller and darker, while the elliptical corpuscles of the lower vertebrata also become rounder : then, in connection with some blood plasma, they become aggregated into small round heaps ; which heaps, by the appearance of an interior nucleus and of an outer meinbrane, experience a transition into spherical cells containing blood corpuscles. These are from 5 to 15-1000ths of a line in size, and contain from 1 to 20 blood corpuscles (figs. 531. 1. 532. 1.) During this time the blood cor puscles are continually diminishing in size, and, assuming a golden yellow, brownish-red, or dark colour, they undergo, either imme diately or after a previous dissolution, a com plete transition into pigment granules. So that these cells themselves are changed into pigmentary granule-cells ; and, finally, by a gra dual loss of colour of their granules, they form themselves into completely colourless cells (figs. 531. 3. 532. 4.).
In respect of the more special circumstances of this process, it is first necessary to con sider the commencement of the cells de scribed, and their changes, somewhat more in detail. As regards the first of these, it is certain that the cells containing the blood corpuscles do not commence directly around a nucleus, but by the circumposition of a membrane around a heap of coagulated blood: in the same way, to wit, that the so-called inflammatory globules of Gluge in certain cases change themselves to cells ; or that by which the smaller globules of fission of the yolk form themselves into vesicles. On the other hand, it remains doubtful whether the nuclei which are seen in these cells are there before the formation of the membrane, or whether they- only begin as supplementary to it. If the former be the case, one might add that, in the extravasated or clotted blood of the spleen, nuclei arise in consequence of the commencing organization, each of which then, like the nuclei in the fission of the yolk, surrounds itself with a part of the blood (plasma and globules), and, finally, con ditionates the development of a membrane on the surface of the sphere thus commenced. Or one might regard the formation of spheres consisting of some blood plasma and blood globules as the primary phenomenon ; and that then a nucleus begins in each sphere; and that, finally, a membrane is thrown around these. In corroboration of this opinion, Hasse and myself* have observed in the pigeon the occurrence of inflammatory globules, which are without nuclei or menibranes, but contain blood globules ; and to this may be added, that in the splenic extravasations blood cor puscles are often grouped together in heaps without being contained in cells. Be this as it may, in any case thus much is certain, that as soon as the cells with their included blood globules are visible, the nuclei are never absent ; and this fact, taken in conjunction with what is already known of the import of nuclei in the process of cell development, speaks strongly for their formation preceding that of the membrane of the said cells.
These cells containing blood corpuscles behave themselves so far alike in all creatures, that their blood corpuscles by degrees dis appear and fall to the ground ; and, ulti mately, they all seem to be converted into colourless cells, although the methods by which this change occurs are different in different animals ; whence it will be well to go through them one by one.
a. In mammals the cells with unchanged blood corpuscles are not very easily seen, on account of the small size of the latter, and the facility with which they lose their colour ; yet one can easily get a sight of them, pro vided the examination be made at the right time, and the application of water forborne. 1 have seen them plainly in man, the rabbit (fig. 531. 1.), guinea-pig, sheep, calf, and dog ; and have found that in these creatures the number of the included blood globules is from 1 to 12, on an average from 2 to 6, and the size of the cells from 5 to 16-1000ths of a line; while their vesicular nuclei have a length of 36-10000ths, and a breadth of 28-10000ths of a line. By the shrinking up and falling to pieces of the blood globules, which immediately renders them darker in colour, coloured granule-cells begin from these cells. They are of a golden yellow, or rusty or brownish yellow, or even blackish colour ( fig. 531. 2.), and gradually experience a trans ition into cells, with slightly coloured, more numerous, and smaller .ranules; and, finally, they take the form oe'altogether colourless cells, part of which are even poor in granules (fig. 531. 3.). In man, the rabbit, and the guinea-pig were found, besides the cells just described, free granules and heaps of granules, of a golden.yellow, brown or blackish colour; together vvith altered blood globules, con cerning which it seemed to me very probable that they were originally free, and were never included in cells. In other vertebrata, as in the hedge-hog, the cat, and the bat (Vespertilio myotis and pipistrellus), the cells with the unchanged blood globules were not observed, although all other stages, from the golden yellow to the altogether colourless granule cells, were seen. Finally, in others, as in the horse and ass, were seen uncommonly nu merous, diminished, and highly coloured blood globules, both isolated and aggregated ; and the metamorphoses of these into golden, brown, and blackish-yellow heaps of granules, although no definite indication of cell struc ture could be detected around these heaps.