194 PROCEEDINGS OF THE ACADEMY OF [Feb., 



tion to form bivalent ones in the first spermatocytes, and all the de- 

 scribers except McClung agree that in the first maturation mitosis they 

 always divide reductionally. No set of chromosomal structm-es is 

 better adapted than such heterochromosomes to prove prereduction: 

 there are two in the spermatogonium, which unite to form a bivalent 

 one in the spermatocyte, and the separation of the univalent halves of 

 the latter in the first mitosis is settled beyond any question of doubt 

 for almost all the cases — for all the cases in which they can be recog- 

 nized by peculiarities of form or size during this mitosis. Never in the 

 spermatocytes do they take on the puzzling forms of rings and crosses 

 which have misled so many good observers in the argument for post- 

 reduction. And it is significant that Gross (1904) shows the bivalent 

 heterochromosome of Syromastes is prereductional in its division, and 

 only by very indirect evidence attempts to show that the ordinary 

 chromosomes divide postreductionally. As to the behavior of this 

 kind of heterochromosome in the second maturation, for most of the 

 species nothing positive could be decided ; for other cases it has been 

 shown that in some cases it divides in the second mitosis (probably 

 equationally), as in Euchistus, Harmostes, Protenor, CEdancola, Syro- 

 mastes, Syrbula, Lycosa and Hippiscus, while it does not divide in this 

 second mitosis in Anasa, Pyrrhocoris and Anax. 



Secondly, as to the division of the heterochromosomes that occur 

 singly in the spermatogonia, and so undergo no conjugation in the 

 spermatocytes. Those of Orphania (de Sinety), Gryllus (de Sinety 

 and Baumgartner), Brachystola (Sutton), and Xiphidium (McClung 

 do not divide in the first maturation mitosis, but do so in the second. 

 Hence here again is prereduction: a whole chromosome passing imdi- 

 vided into one of the second spermatocytes, in the very mitosis which 

 all these observers consider to be equational ! The exceptional case is 

 the unpaired heterochromosome of Protenor ("chromosome x"), which 

 I described (19016) as dividing transversely in the first mitosis, but 

 not dividing in the second. I have recently gone over these old prepa- 

 rations with great care, and find nothing incorrect in my original 

 description. 



Thirdly, in regard to the divisions of the odd chromosomes of CEdan- 

 cola, Harmostes and Alydus, which occur singly in the spermatogonia 

 but are not heterochromosomes. In my original description (19016) 

 I did not determine their behavior positively in Alydus and Harmostes, 

 beyond showing that they do not divide in one of the mitoses, I 

 have recently studied them again, and find that in all these forms they 

 di\nde in the first maturation mitosis but not in the second, just as is 



