264 THE MOSAIC STAGE OF DIFFERENTIATION 



thence into the blood-stream. After being found in all parts of the 

 body for a considerable period, they become localised during a few 

 hours in the site of the future gonad-rudiment. Presumably the 

 gonad-field, once determined, attracts the germ-cells chemically. 



By various lines of evidence it has been shown that a gonad- 

 field is determined and will differentiate into a gonad with typical 

 sex-cords even in the total absence of primordial germ-cells, and 

 that conversely in chorio-allantoic grafts germ-cells may be present 

 in considerable numbers without giving rise to a gonad. There is, 

 however, evidence that the germ-cells can induce some degree of 

 early gonad-differentiation upon peritoneum which would normally 

 never give rise to germinal epithelium. It is thus probable that the 

 germ-cells are necessary for, or at least normally assist in, the 

 process of gonad-differentiation, but that this effect can only be 

 exerted within a "gonad-field" region of the dorsal coelomic 

 epithelium, whose potencies are highest in the presumptive gonad- 

 region. 



In passing, it should be mentioned that, in some mammals at 

 least, germ-cells appear to arise in situ at comparatively late stages 

 in the already differentiated germinal epithelium. Here early gonad- 

 differentiation cannot be dependent in any degree upon the presence 

 of germ-cells. 



The transition between a state of aiTairs in which an early deter- 

 mination of germ-cell tissue occurs in the endoderm and that in 

 which a late determination occurs in the germinal epithelium is not 

 easy to envisage, but we have at least analogies with the determina- 

 tion of other organs, i.e. the lens (p. 189), or the limbs of Amphibia, 

 which in Urodeles are differentiated very early and are independent 

 of thyroid action, while in Anura they differentiate later and will not 

 display full growth in the absence of a certain concentration of 

 thyroid hormone. Further work is needed to elucidate this point. 



The ovaries of birds and of monotremes are of course asym- 

 metrical, that on the left being large and functional, that on the 

 right reduced, and non-functional. It is interesting to find in the 

 bird that this difference is determined from the outset of differen- 

 tiation of gonad-rudiment not as yet showing any sign of sexual 

 differentiation. When grafted on to the chorio-allantois of other 

 embryos, indifferent gonads of either side may differentiate into 



