ABNORMAL MITOSES IN SPERMATOGENESIS. 189 



squirrel (Citcllus}. In the last-mentioned case, the spermatids 

 also are multinucleate, which would seem to preclude the pos- 

 sibility of their forming normal sperms a point in which this case 

 is quite different from the one described in Lo.ro. 



The second instance of abnormal divisions falling under Case 

 2 presents in many ways the most interesting features of all. It 

 is clear that aside from the method just described of multiplying 

 chromosome numbers by cell fusion, the same result might be 

 produced by some failure in spermatogonial division, such that 

 the daughter chromosomes would be incorporated in a single 

 nucleus, while the extra set of centrioles in some way became 

 suppressed or lost. Such might conceivably have been the genesis 

 of the case described by Morgan ('15), and of various triploid 

 and tetraploid 'mutations.' The case now to be described seems 

 to belong in this category. It was found in a specimen of En- 

 schistus v'ariolarius P. B. (Family Pentatomidoe), which had been 

 collected in the fall of 1918 by Professor Wilson (without of 

 course, any suspicion that it was abnormal), and kept in the lab- 

 oratory during the winter. Upon my return to Columbia in 

 February, 1919, this bug, together with other specimens of Eu- 

 schistus, was very kindly turned over to me by Professor Wilson. 

 The abnormality was found in only one specimen (No. 94) (of 

 those sectioned), and in only one testis of that specimen. The 

 other testis was normal in respect to this particular irregularity, 

 though both presented other abnormal features. Whether the 

 stay in the laboratory under somewhat unfavorable conditions was 

 in any way responsible for the abnormalities is by no means clear ; 

 but from the fundamental nature of the particular abnormality 

 here to be considered, it would seem more probable that it arose 

 during one of the earlier instars and was quite uninfluenced by 

 the abnormal environment of the adult. The material happens 

 to have been fixed in Benda's Flemming (with the omission of 

 the mordanting customary in the Benda method itself), and stained 

 in Fe-hematoxylin. This method, intended primarily for the 

 demonstration of the mitochondria, is not entirely satisfactory for 

 the study of chromatic features ; but the preparations proved en- 

 tirely adequate for the study of most of the necessary details. 



This particular abnormality consists of a tetraploid chromosome 



