The close similarity, which in early stages amounts practically to an identity, between the primordial germ-cells and the cells of the ento derm (especially the yolk-sac entoderm) need not, and almost certainly does not, imply a derivation of germ-cells from entoderm-cells. In the first place, the germ-cells can always be distinguished from the ento dermal cells (in spite of structural and tinctorial similarity between the cells) by the rounded (spherical or oval) form of the germ-cells. The cells of the yolk-sac entoderm, which they resemble most closely, are of irregular shape. Occasionally a germ-cell may become wedged in between the entoderm-cells and so forced into very intimate cyto plasmic relationship and into irregular form, in which condition it may be impossible to distinguish between the two; but the resemblance signifies simply, most probably, a similar low grade of differentiation from the original blastomeres. The germ-cells are therefore not derived from the entoderm-cells by differentiation. These two types of cells are similar because they are practically similarly undifferen tiated, both containing a large, finely granular, pale nucleus and a large cytoplasmic content of yolk-globules. The germ-cells simply remain dispersed among the entodermal-cells after the segmentation stages and from here migrate via the area pellucida to the mesentery and thence to the gonads. The germ-cells here, as in certain invertebrates (e. g., Ascaris, Boveri) most probably have been held apart from the soma-cells since an early segmentation stage and then, from widely scattered areas, have migrated through progressively more sharply segregated foci to the gonads, maintaining thus a continuous germ-cell cycle from blastomeres (perhaps a particular portion of the fertilized egg, as, for example, the pole disk of certain Diptera, Hegner) to the gonocytes of the definitive genital glands.
It remains to discuss the mode of the extravascular migration—i. e., whether active (amceboid) or passive, or both. A passive migration is strongly urged by von Berenberg-Gossler. He thinks the germ-cells are passed from the lateral to the medial entoderm, and from the lateral visceral mesoderm to the mesentery, and then across the ccelomic angle by the mechanical process of unequal growth. Growth is no doubt a factor in the shifting. But the chief factors inhere in the cells them selves, which (judged from their shape and the manner in which the mesenchymal cells are compressed in the direction of a path towards the root of the mesentery) is an active amceboid movement of the cells. The cells evidently move medially both within the endoderm and in the mesoderm. Such as lie medially in the entoderm become inclosed within the hind-gut, whence the majority later migrate into the mesen tery.
Only two germ-cells were found at any stage within any blood channels. But no confusion between blood-cells (hemoblasts) and germ-cells seems possible. The germ-cells are at all stages larger than the largest granulocyte (eosinophil, whose granules also stain black in iron-hematoxylin) and the largest hemoblast, the character and staining capacity of whose nucleus is very different. There is therefore
no room for doubt that in chick (Swift) and duck (von Berenberg Gossler) the germ-cells actually do migrate (passively) towards the geni tal ridge largely through the blood-vessels. The locus of original segre gation in birds is such as to provide in the blood-channels the most favorable method for the longer portion of the route of migration to the sexual gland.
As Allen has already suggested, the reason for the mesodermal origin of the germ-cells in urodeles and possibly certain other vertebrates may be a relatively early separation of mesoderm from entoderm—that is, the germ-cells of the primitive entoderm may have become separated with the portion which contributed to the splanchnic layer of the mesoderm before they could migrate from the entoderm proper.
In Lacerta von Berenberg-Gossler describes cells comparable to the so-called primordial germ-cells passing from the entoderm to both layers of the mesoderm, and to the dorsal layer of the intermediate cell-mass, where caudally the cells are said to contribute largely to the formation of the WolfEan duct. Von Berenberg-Gossler interprets his results to mean a belated origin of mesoderm from entoderm. He very ingeniously suggests that these " germ-cells " wander also to the sex gland, where they may transform into mesenchymal-cells, and then retransform into germ-cells, but presents no histologic data in support. Von Berenberg-Gossler's results and his interpretation are unique and do not seem to be capable of being brought into harmony with the results obtained in any other form thus far studied, nor with any common interpretation. Von Berenberg-Gossler's interpretation is more or less plausible, but can not be said to be wholly satisfactory. The condition in the lizard may be simply an exaggeration of that in other forms where a certain number of germ-cells go astray during the period of migration.
Since germ-cells can migrate by their own activity, it is not difficult to conceive of their presence anywhere, even in the ectoderm, and certainly not in the developing Wolffian duct, where, unless they retrace their course towards the genital gland, they may undergo degeneration (or possibly lie dormant until a favorable influence may excite to the production of a tumor or a teratoma), but that they should contribute to the formation of the Wolflian duct is thus far an anomalous phenomenon.
When one considers in common conditions in the chick, lizard, toads and frogs, salamanders, and turtles, it becomes clear that there is great variation in the manner in which the comparable elements (the "primordial germ-cells") arise, migrate to the sexual gland, and develop. Underlying these discrepant phenomena, however, there appears a fundamental harmony, namely, an early segregation of germ-cells from somatic cells, a germinal path from blastomeres to genital cells of the genital gland, probably from blastomeres to sperm and ova; a migration from the entoderm (or its derivative, the meso derm, in urodeles and lamprey), through the splanchnic layer of the mesoderm to the mesentery and thence to the gonads (with possible aberrant migrants anywhere).