196 



ROBERT H. BOWEN. 



reduction, or from a tetraploid auxocyte after normal reduction 

 has occurred. 



My observations indicate that the requisite conditions for the 

 origin of the triploid, and perhaps tetraploid, individuals may be 

 fulfilled by a combination of the possibilities suggested by the 

 hypotheses outlined above. Thus, it seems certain in the Eu- 

 schistus case that a suspended (or otherwise abnormal 1 ) mitosis 

 has occurred during some early division in the germ cells, giving 

 rise to a tetraploid condition confined to the descendants of the 

 particular cell involved. If, as seems probable, each cyst of 

 spermatogonia is derived by the repeated division of a single 

 cell derived from an early germ cell, the occurrence of the tetra- 

 ploid cysts would be accounted for, and also the diploid cysts 

 produced concurrently from the early germ cells of normal con- 

 stitution. The spermatocytes descended from the tetraploid cells 

 have undergone a reduction in the proper sense of the term, but 

 the number of chromosomes in the resultant gametes is undi- 

 minished ("unreduced") as compared with the number in the 

 normal (haploid) ones. The fusion of such a diploid gamete 

 with a normal one would obviously produce a triploid individual. 

 An explanation along these lines seems to me preferable to the 

 one sometimes given which explains triploidy as due to the union 

 of a normal and an unreduced (properly speaking) gamete; for 

 while such unreduced gametes have been supposed to arise (in 

 Hemiptera) through the suppression of one (or both) of the 

 maturation divisions, they always contain an abnormal number 

 of centrioles (as Paulmier, '99, long ago showed), and the weight 

 of evidence is against the probability of their functioning nor- 

 mally in fertilization. 



The origin of tetraploid individuals by this method would 

 depend upon an exceedingly rare coincidence, and, as a matter of 

 fact, tetraploidy in animals seems to be a rather rare phenomenon. 



1 The case of Lo.va suggests the possibility that a tetraploid germ cell 

 might arise from the fusion of two normal ones. By analogy such a fusion 

 product might well contain the normal centriole content, as observed in 

 Euscliistus. The usual explanations advanced to account for the doubling 

 of chromosomes all fail to account for the observed normal number of cen- 

 trioles an important part of the story, which has not yet received adequate 

 attention. In this connection it is of interest to note that centrioles are 

 generally absent in the higher plants. 



