1905.] NATURAL SCIENCES OF PHILADELPHIA. 187 



the strongest argument yet given for postreduction; he is the sohtary 

 worker on the spermatogenesis of Hemiptera who has taken the post- 

 reductional view, and does it from a supposed secondary change in 

 chromosomal form. Yet, strangely enough, he describes a prereduc- 

 tional division of the bivalent chromatin nucleolus, the only chromo- 

 some which is not said to pass through the stage of a cross! Its two 

 whole univalent components become separated from each other in the 

 first maturation mitosis. In his object the bivalent chromos®mes are 

 in some stages usually little longer than broad ; they approach in some 

 conditions more nearly the form of a cross than in any Hemipteron 

 which I have studied. Had Gross taken up my old object, Euchistus, 

 he would have found that X-shaped chromosomes do not occur at all, 

 or only very rarely, that the phenomena there are accordingly simpler 

 and more explicable than in Syromastes, and that intermediate forms 

 between a chromosome elongated in one direction and one stretched 

 out in another do not occur. Finally Gross admits that these forms 

 admit of another interpretation: "Man konnte mir entgegenhalten, 

 dass der von mir aus den Thatsachen erschlosene Modus der Tetraden- 

 bildung auf einer wdllkiirlichen, durch nichts bewiesenen Annahme 



beriihre Sichere Anhaltspunkte dafiir, nach welcher Richtung 



die Halften der Kreuzc aus einander weichen, lassen sich aus den beo. 

 bachteten Figuren nicht entnehmen." I fully agree with him in this- 

 But when he states, "Dasselbe gilt aber auch von der bis jetzt allge- 

 mein angenommenen Bildungsweise," he makes an error, for in some 

 cases of spermatogenesis cross-shaped chromosomes do not occur, and 

 that is so in Euchistus, and for such forms no voluntary assumptions are 

 necessary. Gross' work appears very accurate, and I criticise only his 

 interpretation of the crosses as intermediate forms. Evidently he is 

 considerably influenced by Hacker's latest views. The same general 

 criticism may be made of the work of Sutton (1902). 



When we review all this work supposed to prove a postreduction, 

 we find it based upon an incomplete series of stages, or upon forms 

 with minute chromosomes of very diverse form, or upon such as have 

 chromosomes in the form of rings and crosses. Every one will admit 

 that chromosomes of such shapes are the most difficult to interpret : a 

 tetrad with four parts of approximately equal size — where in it can we 

 say lies the plane of the longitudinal split and where the line separating 

 two univalent chromosomes? Just upon such chromosome forms is 

 much of the postreduction argument based. The correct, because only 

 decisive, method is not to reason from such forms, not to argue unneces- 

 sarily for a change in axis, but to explain such chromosome formation 



