Ill DIFFERENTIATION 107 



tion continues in later divisions until ultimately about sixteen 

 or seventeen chromosomes in all are got rid of. The germ- 

 number in Strongylocentrotus is eighteen, and among these is 

 a long hook-shaped chromosome, which is always found in the 

 hybrids ; the germ-number in Sphaerechinus is twenty, and 

 amongst these are two very long chromosomes ; these are 

 never found in the hybrids. 



In the larva reared from the disperm egg there should 

 therefore be 18 -f 20 + 20 chromosomes, but the same elimina- 

 tion occurring, there are only twenty-six, while the monosperm 

 larva has twenty-two. In other words, the disperm larva has 

 lost thirty-two while the monosperm has lost sixteen, the 

 disperm has retained 18 + 4 + 4, while the monosperm has 

 retained 18 + 4, Hence it is the paternal chromosomes which 

 have been eliminated, all but four. In the cross Echinus a* x 

 Sphaerechinus ? there is no elimination of chromosomes, and 

 the larva is intermediate. In the reciprocal cross the elimi- 

 nation occurs, but the paternity of the lost chromosomes has 

 not been demonstrated, though it is known that the charac- 

 teristic long hook and horse-shoe of Echinus remain. 



In the cross Strongylocentrotus $ (or Echinus $ ) x 

 Arbacia a*, the larva is of the maternal type, and there is 

 elimination of chromosomes. This, however, does not occur 

 until the blastula stage is reached. It is not possible to be 

 certain that the eliminated elements are paternal. In the 

 reciprocal cross there is also elimination but the pluteus stage 

 is never reached. 



The discussions of the foregoing section may now be very 

 shortly summarized. 



1. Experiment shows that the removal of certain parts of 

 the cytoplasm of the ovum entails the absence or at least the 

 defective development of certain organs of the embryo or 

 larva. Hence there are in the cytoplasm certain material 

 factors on which the formation of certain characters depends. 

 These characters are part of the total inheritance. 



2. Every visible substance in the cytoplasm is not, however, 

 necessarily such an organ-forming substance as experiments 

 with the centrifuge demonstrate. 



