THE MOSAIC STAGE OF DIFFERENTIATION 263 



notably partial twinning. Thirdly, teratomorphic and malignant 

 effects. It is interesting that, as Dr Waddington has pointed out 

 in conversation, carcinogenic compounds are certainly related to 

 oestrin, and are probably to the chemical substance responsible 

 for organizing (p. 154). So our three effects, concerned with 

 sex, wath organiser abnormality (twinning), and with malignancy, 

 may conceivably all be related to one fundamental process af- 

 fecting substances of this type. 



It is interesting that keeping the eggs for some time before fer- 

 tilisation in conditions of relative lack of w^ater (hypertonic salt 

 solutions ; keeping in air with a minimum of moisture) leads to a 

 large preponderance of females.^ Unfortunately no embryological 

 study of this case has been made, but it too is evidence that condi- 

 tions in the egg-cytoplasm during the earliest stages of development 

 may modify morphogenesis at later stages : since sex-differentiation 

 provides two alternative methods of morphogenesis in which the 

 result is determined by a balance of two competing factors, w^e 

 should expect to find in it the best indicator for such effects. 



In Amniotes, the embryology of the gonad is not so simple. In 

 general, however, we may say that the distinction between mascu- 

 linising medulla and feminising cortex is maintained. The medulla 

 is largely formed by the primary sex-cords w^hich migrate inwards 

 from the germinal epithelium; in the male these form the testis 

 tubules, in the female they become inhibited and persist in modified 

 form. In the male, only one set of sex-cords is formed. In the 

 female, however, the cortex enlarges to produce a second set, which 

 gives rise to the main ovarian structures, in association with which 

 the female germ-cells differentiate. In the male, on the other hand, 

 the cortex (germinal epithelium) becomes reduced to a mere peri- 

 toneal epithelium soon after the formation of the primary sex-cords. 



The chick is here the best-investigated type. In the chick the 

 germ-cells are formed in the extra-embryonic endoderm, in a 

 crescent-shaped area of the blastoderm antero-lateral to the embryo. 

 Embryos can be castrated by excision of this area, or by ultra- 

 violet^ or X-rays,^ proving the mosaic determination of the pri- 

 mordial germ-cell tissue. Normally these cells appear to be at- 

 tracted into the mesoderm when it invades the germ-cell field, and 



^ King, 1912. - Reagan, 1916; Benoit, 1930. * Danchakoff, 1933. 



