72 THOS. H. MONTGOMERY jr., 



threads, as is stated and figured, each of them can have only the 

 value of half a chromosome, and hence there is no doubling of the 

 normal number of chromosomes, as Wilcox assumes. This is a case 

 where the morphological value to be ascribed to a chromosome must 

 remain a matter of individual opinion, unless we start with a fixed 

 definition for the term chromosome. Wilson ('96) also concluded there 

 must be only 12 chromosomes in the 1st spermatocyte of Caloptenus, 

 but reached his conclusion by throwing doubt upon the accuracy of 

 Wilcox's observations as to the absence of any longitudinal splitting 

 of the chromosomes, which objection has been shown by Wilcox to 

 be baseless. The 12 dumbbell-shaped chromosomes apply themselves 

 together in pairs, and so are formed 6 pairs of chromosomes (not 6 

 groups of 4 chromosomes each, as assumed by Wilcox). For let us 

 compare the chromosomes in Caloptenus with those in Pentafoma. In 

 the latter occur in the spermatocytes of the first order 7 dumbbell- 

 shaped chromosomes (half the normal number); in these we find a 

 segregation of the chromatin at two poles, though, as a rule, without 

 a line of interruption between the two masses of chromatin. But 

 sometimes in Pentatoma the segregation of the chromatin in a chromo- 

 some may proceed so far, before the nuclear membrane has dis- 

 appeared, that the chromosome is formed of two masses of chromatin 

 connected together only by a narrow bridge of linin. This then would 

 have some resemblance to the dumbbell-shaped chromosomes of Calo- 

 ptenus; but in both forms the dumbbell-shaped element had the same 

 mode of origin, and hence must be still considered a single chromo- 

 some, since the connecting bridge of lining docs not become broken 

 until metakinesis. My standpoint of the value of the dumbbell-shaped 

 elements in Caloptenus is upheld by the observation of Wilcox in 

 regard to the second, less frequent mode of origin of them, in which 

 the spirem breaks into 12 chromosomes, which do not show transverse 

 constrictions until they have become arranged in pairs. Thus, accord- 

 ing to my definition of a chromosome, the normal number of chromo- 

 somes appears in the 1st spermatocytes of Caloptenus. 



Now Wilcox finds that every two such dumbbell-shaped elements 

 associate together so as to form a closed ring, by the apposition of 

 the two ends of the chromosomes of a pair ; thus each ring represents 

 4 small, spherical masses of chromatin connected together so as to 

 form a ring or square. Now since he regards each ring as equivalent 

 to 4 chromosomes, he considers that the spermatocytic divisions have 

 the value of transverse divisions of the chromosomes, and that each 



