1905.] NATURAL SCIENCES OF PHILADELPHIA. 195 



the case with the unpaired heterochromosome of Protenor and what 

 Gross (1904) has called the "accessory chromosome" in Syromastes (to 

 which we shall reciir). They do not appear bivalent in the first sper- 

 matocytes ; and whether their division in the first maturation mitosis 

 is transverse or parallel to their long axis was not determined on account 

 of their nearly spherical form. 



Now to him who has had the patience to follow tWs account, which 

 gives only a brief statement of some of the results of previously 

 detailed observations, the occurrence and behavior of the two kinds of 

 heterochromosomes and of the odd ordinarj^ chromosomes may well 

 seem difficult to reconcile. But there is nevertheless a general con- 

 formity of process here, which has not been elucidated heretofore. 

 Whenever the heterochromosomes occur in pairs in the spermatogonia 

 they always conjugate to form bivalent ones in the first spermatocytes, 

 and their univalent components become separated in the first matura- 

 tion mitosis, i.e., divide prereductionally. This is strictly in confirma- 

 tion with the doctrine we have tried to lay down in this paper, that the 

 separation of entire univalent chromosomes, i.e., their reduction in 

 number, is always accomplished in the first mitosis. At the same time 

 we have to bear in mind that there is no evidence that chromosomes 

 divide in different ways in the first maturation mitosis, some equation- 

 ally and some reductionally ; it is very probable that does not happen, 

 and indeed until proof is brought to the contrary we are justified in 

 maintaining that it does not occur. This is an important premise in 

 interpreting the divisions of the heterochromosomes and ordinary 

 chromosomes that occur singly in the spermatogonia. Now in the 

 Orthoptera (Orphania, Gryllus, Xiphidium, Brachystola) the hetero- 

 chromosome is single in the spermatogonia; single, therefore, in the 

 spermatocytes, it does not divide in the first maturation mitosis, but 

 does in the second. Because it does not divide in the first mitosis it 

 must be either univalent or else already in the spermatogonia be com- 

 posed of two so firmly united that they cannot be divided in the reduc- 

 tion mitosis; its division in the second mitosis must be equational, 

 and all the descriptions show this to be so. Now in Protenor the case 

 is reversed; the single heterochromosome divides in the first mitosis, 

 but not in the second, exactly like the odd ordinary chromosomes of 

 the Hemiptera, but apparently the reverse of the single heterochromo- 

 somes of Orthoptera. Since this heterochromosome of Protenor and 

 the odd ordinary chromosomes of three other Hcmipteran species 

 divide during the reduction mitosis, these chromosomes must be 

 already bivalent within the spermatogonium — the single one there be 



