MULTIPLE CHKOMOSOMES 551 



chance relations between masses of substance in different 

 'phases,' but through a knowledge of the history of certain 

 individually recognizable chromosomes traced through a large 

 group of animals. Worthy of note is also the fact that in such 

 cases as H. viridis the range of variation is strictly defined and 

 falls within the Hmits set by the organization estabhshed in the 

 family, except in the occasional instance of supernumerary 

 chromosomes — two instances in the thirty-eight individuals 

 studied.'* 



It is only upon the basis of entire identity of all chromo- 

 somes in a complex and of their temporary character that the 

 Itahan author has any argument at all. Building upon this he 

 seizes upon every reported instance of difference of chromosome 

 numbers as a support for his thesis. Every descriptive fact is 

 labeled a 'sub hypothesis,' and so many of these are secured in 

 this way as to convince him that the main hj^othesis of chro- 

 mosome individuahty has no standing. Curiously enough, he 

 conceives a great importance for the fact that one explanation 

 for all the reported cases of numerical variation is not sufficient. 

 On the contrary, he says, there are so many ways of accounting 

 for such variation that there can be no constancy and no indi- 

 viduality. Perhaps in no other way is the quality of his argu- 

 ment better indicated than here. The fact that an error in 

 observation has been made, and even admitted, is only an indi- 

 cation of variation; when it is shown that a reported difference 

 in number is due to the inclusion of more than one species in a 

 study this is a subhypothesis weakening the main one ; the pro- 

 duction, by hybridization, of chromosome numbers different 

 from those of the parents is accounted as an example of varia- 

 tion, as is the case of asymmetrical mitoses in neoplasms. The 



* If a chromosome is but a chromosome without character or distinction, there 

 is in this case a violation of the rule stated. But if the nature of the element 

 be significant, the rule holds, for the entire history of the supernumerary chromo- 

 some demonstrates that it is of a different order from the euchromosomes. It 

 may in many ways depart from the history of typical chromosomes and finally 

 end in complete elimination. Arguments, such as Delia Valle's, require that 

 such a structure be considered identical with all other bodies called chromo- 

 somes, without in any way regarding its individual history of ultimate extinction. 



