The germ-cells. 663 



of 6 to 20 mm. They also vary in size from 0.02 to 0.036 uj) to 0.056 mm. 

 The smaller of them of which there are many drawings in the plates 

 (Figs. 21, 23 to 28, 30, 31) are of exactly the size and appearance 

 of normal germ-cells. The larger ones (Figs. <S, 13, 18, 29, 36) only 

 diÖer from the latter in size. In all other respects — unless they 

 be multinuclear, and, therefore, in degeneration — they are quite like 

 normal germ-cells. And this is true of them, wherever they may be. 

 Their size, however, is always such that one, two, three, or, at most, 

 four mitoses would bring them down with great accuracy to the 

 average size of an ordinary germ-cell. If size were to be a bar to 

 their germ-nature, it would also prohibit any genetic connection between 

 the first-formed leucocytes and the later and smaller ones, which un- 

 doubtedly arise from them. 



And in fact, as a reductio ad absurdum, the first cells of epiblast, 

 mesoblast, and hypoblast could not be connected with the later ones ; 

 because, as even the present drawings show, they are much larger 

 than the latter. 



Apart from the resemblances between these "megaspheres" and 

 germ-cells, there are also the facts of similar features in -their de- 

 generation. We cannot account for the small number of germ-cells 

 in early embryos, or their absence from the mesentery, where later 

 on they are so abundant, without the inclusion of the wandering 

 "megaspheres" among the germ-cells. Moreover, there is such an 

 unbroken transition from the ordinary germ -cells of 

 my embryos to the largest of the "megaspheres", and 

 so many of the latter agree so absolutely with ordinary 

 germ-cells in every respect except in position, that it 

 is quite impossible to draw any line between them. In 

 the normal position, in the germinal nidus, one also encounters ab- 

 normally large germ-cells or "megaspheres". 



To sum up : the interpretation of the "megaspheres" as germ- 

 cells, or, if large, as the forerunners of such, accounts for them 

 in such a manner as to make any further explanation a super- 

 fluity. It is required, in order to avoid the insuperable difficulty of 

 attempting to draw a hard and fast line, where such a thing has no 

 shadow of existence, either in size, characters of the two sets of cells, 

 or their positions in the embryo. It is imperative, because without 

 the "megaspheres" the missing germ-cells of earlier embryos cannot 

 be accounted for, in default of mitoses among the few normal and 

 ordinary germ-cells in any embryo of 6 mm. The degenerative 



