46 SCIENCE PROGRESS 



moth that although a large variety of stimuli lead to an 

 increased proportion of eggs which segment, yet there is no 

 common characteristic such as dehydration which can be 

 regarded as a general cause ; and since Loeb admits that the 

 eggs of some animals may be caused to develop by other 

 means — though abnormally — it is perhaps premature to regard 

 loss of water as the invariable stimulus required for their 

 development. 



Hitherto we have considered only what may be described 

 as the natural history of parthenogenesis, the conditions under 

 which it is found in various groups, and how it is related to 

 bisexual reproduction. But one of the chief interests of a com- 

 parison of these two forms of reproduction is in the differences 

 in the maturation of the eggs in the two cases. In eggs which 

 are to be fertilised it is invariably found that the chromosome 

 number is reduced to one-half that which is normal in the 

 somatic cells of the species, and this almost always happens 

 by a preliminary pairing of the chromosomes (or some 

 equivalent process) followed by two nuclear divisions, one 

 of which separates the members of each pair into different 

 daughter nuclei, and the other splits each into two similar 

 halves. These two divisions give rise to the first and second 

 polar bodies, which, with very rare exceptions, are formed by 

 eggs requiring fertilisation. A similar process occurs in the 

 maturation of the spermatozoa, a single unreduced cell giving 

 rise by two successive divisions to four cells, each containing 

 half the normal number of chromosomes. 



The actual process of " chromosome reduction " appears to 

 vary somewhat in different cases, and there has been much 

 disagreement among authorities as to how it typically takes 

 place. Perhaps we may regard the phenomena described by 

 Farmer and Moore 1 as most generally typical. At the beginning 

 of the " meiotic phase " (the stage during which the reduction 

 divisions take place), chromatic loops appear in the nucleus, 

 in number half as many as the normal number of chromosomes 

 in ordinary cells. These loops frequently show signs of a 

 longitudinal split. They become shorter and thicker, and con- 

 tract into characteristic ring or loop shapes which arrange 

 themselves on the spindle in such a way as to be divided 

 transversely instead of longitudinally. Even at this stage the 



1 Farmer and Moore, Q.J. M.S. vol. xlviii. 1905, p. 489. 



