74 EDMUND B. WILSON 



explanation of the facts. It will be evident to readers of my for- 

 mer papers that I am fully prepared to accept Boveri's conclusions; 

 but there is one very important fact, finally established by the 

 present paper, that must be clearly recognized. If we assume that 

 different factors of heredity are in some sense unequally distri- 

 buted among the chromosomes, we need feel no surprise that the 

 duplication of one or more of the ordinary chromosomes should 

 produce no perceptible qualitative effect upon development. But 

 it is very surprising that no visible effect should be produced by the 

 removal of a particular chromosome that has no duplicate to take 

 its place. In preceding papers I have called attention to the sing- 

 ular fact that Montgomery's material of M. terminalis differs con- 

 sistently from my own in the lack of the small idiochromosome or 

 u Y-element" (see Wilson '09a, for this term); but the possibility 

 of two distinct species or races having been confused could not be 

 absolutely excluded. In the present case, however, no doubt can 

 exist, since the two original specimens of M. femoratus from Miami, 

 Florida, are in my cabinet, and in perfect condition for identifica- 

 tion. One of these, as already stated, contains both the small 

 idiochromosome and an additional supernumerary, while both are 

 lacking in the other individual; yet the two individuals seem to be 

 otherwise in every essential respect identical. All doubt is thus 

 removed that the small idiochromosome or Y-element, which forms 

 the synaptic mate of the accessory chromosome or X-element, 

 may be present in some individuals and absent in others of the 

 same species, without the appearance of any corresponding dif- 

 ferences in the sexual or other characters as far as shown in the 

 external morphology of the animal. 9 This chromosome, as shown 



9 It will be seen from this how readily discrepancies regarding the number of 

 chromosomes might arise between different observers working on the same species. 

 It might seem that we have here a simple and plausible explanation of the contra- 

 dictions that have arisen in the case of Anasa tristis; for we might assume that the 

 diploid number is 21 in some individuals of this species and in others 22 ; and a simi- 

 lar explanation has in fact already been adopted by more than one recent writer 

 (cf. Delia Valle: '09, Buchner: '096). 



I am not myself able to take this view of this particular case for several reasons. 

 In the first place, if there be individuals of this species that have 22 spermatogonial 

 chromosomes, as maintained by Foot and Strobell ('07) we should expect to see 



