The spermatogenesis of Peripatus (Peripatopsis) balfouri. 279 



modified to form the basis for comparative generalisations, and that 

 rather the more conservative elements of the cell, the chromatin and 

 the Uuiu connected with it, should be the object of research. 



I. Material. 



I wish to express here my hearty thanks to my old friend 

 Dr. W. F. PuRCELL of the Museum of Natural History, Cape Town, 

 Africa, for his kindness in sending me the valuable material of Peri- 

 patopsis. Dr. PuRCELL collected the material personally in the vicinity 

 of Cape Town , and identified the species as Peripato2)sis halfouri 

 Sedgwick. In his recent revision of the South African species of 

 Peripatidae, Dr. Purcell (1899) adopts Pocock's genus Peripatopsis 

 for most of the African species of the family, and following Pocock, 

 restricts Peripatus to the American forms. 



The material at my disposal consisted of 32 whole testes (with 

 their seminal vesicles, and proximal portions of their vasa deferentia). 

 A few of these had been fixed for one hour in a concentrated aqueous 

 solution of corrosive sublimate, with 2 % acetic acid; of the others, 

 the majority were fixed in Flemming's osmic-chromic-acetic mixture, 

 undiluted, for 24 hours, others in diluted solution of this fluid for 

 the same length of time. The fixation of all these testes proved to 

 be excellent, but for the cytoplasmic details the osmic mixtures proved 

 the better. Also five ovaries (with the proximal portions of their 

 oviducts) were sent me, and one of these contained an embryo; these 

 had been preserved by the same methods. 



All the ovaries were serially sectioned (sections of 6 fi thickness), 

 and 20 out of the 32 testes. The stains which gave the best results 

 were Heidenhain's iron haematoxyline , and Hermann's safifranine- 

 gentian violet stain. The following were also employed as a control 

 to the others, and in the endeavor to find a good stain for the linin 

 structures ^) : Flemming's triple stain , the Ehrlich-Biondi-Heiden- 

 HAiN stain, Ehrlich's haematoxyline with eosin, Lyon's blue, acid 

 fuchsine in 50 7o alcohol (in various combinations) and Kernschwarz. 



1) I would suggest for the purpose of obtaining a good linin 

 stain and chromatin stain simultaneously, the combination of iron haem- 

 atoxyline and acid fuchsine; careful experiment with this method might 

 produce a strong red stain for the linin, and a blue or black stain for 

 the chromatin; some sharply dififerentiated stain like this is very much 

 needed. 



