It may be added also that my previous study of the early history of the female germ-cells of the starfish Asterias forbesii (1908) had led me to the conclusion that in this form the germ-cells most probably arose by differentiation from the peritoneal epithelium, and so inclined me to a belief in the essential accuracy and general applicability of Waldeyer's original theory. But a steadily augmenting body of cytologic and genetic data would seem to demand a continuous "germ plasm," which demand does not appear to be met by a derivation of germ-cells from already differentiated soma-cells.
Material AND METHODS.
The first essential requisite was a complete series of embryos. The second requisite was a favorable staining technic, such as would clearly differentiate the primordial germ-cells at all stages of their early his tory; which condition was in turn dependent upon the employment of such a fixative as faithfully preserved differential structural charac teristics.
The material embraced in this investigation, selected from a large stock supply, included embryos from the second to the thirty-second day of incubation, distributed, preserved, and stained as follows: This material was collected on Loggerhead Key, Tortugas, Florida, in the summer of 1914, while at the Laboratory of the Department of Marine Biology of the Carnegie Institution of Washington.
Three nests of eggs were available for the earlier parts of the series. Two additional nests contributed to the later parts of the series. The difficulties of removing the early blastoderm from the yolk were not successfully overcome until the time had passed for preserving stages younger than the end of the second day (5-somite stage). However, this stage shows clearly the origin of the germ-cells from among the yolk-sac entoderm, and the beginning of their segregation into two bilateral cords near the lateral margin of the area pellucida in the caudal half of the blastoderm; and it corresponds closely with von Berenberg-Gossler's illustration (fig. 1) of his youngest lizard embryo (5-somite stage).
Two chief reasons led to the selection of the loggerhead-turtle embryo for this investigation: (1) the possibility of obtaining a large and closely graded series of stages; the number of eggs per nest is near 100; development is comparatively slow, the incubation period being about 8 weeks; (2) I had already familiarised myself in previous works with the embryonic blood-cells of certain Chelonia (1913) and also with the male germ-cells in a study of the spermatogenesis of several species of turtles (1914).
The object in using Helly's fluid as a fixative almost exclusively was to preserve the cytoplasmic granules, both albuminoid and lipoid, that is, blood-cell granules and mitochondria, for subsequent differen tial staining with the Giemsa stain and with iron hematoxylin. Paral lel series were prepared with these two stains, with the expectation of facilitating the differentiation between hemoblasts and gameto blasts. This technic was very serviceable, but did not yield quite the fine uniform results hoped for. As concerns the hemoblasts and granu
lar blood-cells (eosinophiles) nothing better could hive been desired; but the cytoplasmic contents of the primordial germ-cells were only indifferently preserved. The fixation period should probably have been prolonged for 3 or 4 days beyond the usual 24 hours, especially in the case of the older embryos. Only very rarely were mitochondria and yolk-globules preserved in the later stages. Especially in the germ-cells were the yolk granules dissolved, leaving thus a greatly vacuolated and distorted cytoplasm. But the Helly's fluid and the subsequent treatment involved in the paraffin technic had a very variable effect upon the yolk-granules at different stages. The same type of cell might contain many or only a few, or no yolk-granules. In the latter case the yolk-content had been entirely dissolved. Germ cells of the 2-day stage of incubation showed many yolk-spherules; by the fourth day no germ-cells contained yolk after this fixation. Obviously the later steps in yolk metabolism were more susceptible after Helly's fixation to the solvent action of the alcohols and oils used in the paraffin technic. The young germ-cells evidently contain less readily soluble yolk than other cells, with the exception of the entodermal cell of the area opaca. In tissues fixed with Helly's fluid and stained with iron-hematoxylin the germ-cells in early stages therefore stand out clearly from among the entoderm-cells of the area pellucida, and accordingly this technic proved to be very favorable for the study of this early, most important, stage.
Some embryos of the eleventh and the twenty-fifth day of incubation were fortunately preserved in Flemming's fluid. It so happens that the 11-day stage of the incubation period is crucial from the viewpoint of the greatest abundance of primordial germ-cells in the area including the closed hind-gut, the mesentery, and the primitive gonad. In these cells mitochondria as well as yolk-granules are well preserved; and so also are the granules of the confusing large blood granulocytes. The germ-cells in the gonads of the 25-day embryo are practically iden tical with those of the 11-day embryo in form and size and in regard to cytoplasmic content. Comparative study of the 11-day embryo fixed and stained in the three different ways—(1) Flemming's fluid and iron-hematoxylin stain; (2) Helly and iron-hematoxylin; (3) Helly and Giemsa—revealed admirably the diagnostic marks between the germ-cells and cells with which they might be confused, i. e., blood granulocytes, yolk-laden hemoblasts, and yolk-laden entodermal cells. These marks once established at this stage, earlier stages could be much more readily identified in the Helly-Giemsa or Helly-iron-hema toxylin material than in the Flemming material, since the most con fusing and obscuring features were here largely eliminated, namely, the very abundant yolk globules and granules.
Descriptive.