The Spermatogenesis in Pentatoina up to the Formation of the Spermatid, 71 



chromatin elemcüts of secoudiirily modified reduction divisions. We 

 may now apply this definition to those cases of reduction divisions 

 which have been more fully and accurately described, i. e. to those 

 cases, where the sexual products have been followed continuously from 

 the spermatogonia (or ovogonia) to the spermatid (or ovum). And 

 since in general there is a marked correspondence between the pro- 

 cesses of ovogenesis and spermatogenesis (first shown by Henking, 

 '90, and IIertwig, '90, independently of one another), these two pro- 

 cesses may be considered together. The following objects have been 

 relatively accurately described, and may be treated in succession : 

 Ascaris^ Copepoda {Cyclops, Diaptomus, Heterocope, Euchaeta, Cantho- 

 camptus, Calanus, Eucalanus, Pleuromma, Anomalocera), Salamandra, 

 Selachii, Calopfenus, Gryllotalpa, Lumhricus, Ophryotrocha, PyrrJio- 

 coris and Pentatoma. The following seem to have been not sufficiently 

 well described, but may be referred to subsequently: AUolohopJiora 

 (Foot), Sttjelopsis (Julin), Pygaera and Sphinx (Platner), Bombyx 

 (Toyama), Diaptomus (Ishikawa), and Mantis (Giardina). 



Ascaris megalocephala hivalens. In the prophase of the first 

 reduction division there is but one transverse segmentation of the 

 spirem thread, and each of the portions then undergoes two longi- 

 tudinal constrictions (Hertwig, '90, Brauer, '93b, Sabaschnikofp, '97). 

 Thus, there being division of the spirem thread into only two parts, 

 there occur here two quadripartite chromosomes (Boveri, '92a, Brauer), 

 and not 8 unipartite chromosomes (Hertwig). 



In Copepoda (Hacker, '95, which contains a summary of his pre- 

 ceding observations ; RtJCKERT, '94, Vom Rath, '95) the spirem thread 

 breaks into half the normal number of chromosomes, each of which 

 undergoes one longitudinal and one transverse constriction in the pro- 

 phase. According to our definition, then, there are here half the normal 

 number of chromosomes, and each of them is quadripartite. 



In Salamandra (Meves, '96) the chromosomes appear in half the 

 normal number in the 1st spermatocyte, and only in the heterotypic 

 generation is there a splitting of the chromosomes before metakinesis. 



In Selachii (Moore, '95) the number and formation of the chromo- 

 somes is essentially the same as in Salamandra. 



In (7rtZop^erms(WiLGOx, '94— 97)the normal number of chromosomes 

 is 12. The spirem thread of the 1st spermatocyte breaks up into 12 seg- 

 ments, and according to my definition there must be accordingly 12 chromo- 

 somes present. Each one of these chromosomes is described as dividing 

 into two, but since the two halves remain connected together by linin 



