The spermatogenesis of Peripatus (Peripatopsis) balfouri. 311 



which is generally assumed to be related to the ancestors of Insects, 

 would throw light upon the phylogenetic origin of the chroraatin 

 nucleolus of Insects, that curious nuclear structure first fully described 

 by me for the spermatocytes of Peniatoma, and shown by me to be 

 probably a metamorphosed chromosome, and so to have no genetic 

 connection with the true nucleoli. Now there are two methods of 

 determining the morphological value of a nucleolar structure: the 

 determination on the basis of its chemical structure, which is however 

 much less reliable than the second mode, namely the determination on 

 the basis of its morphological genesis. The nucleoli of Feripatus resemble 

 the chromatin nucleoli of Pentatoma in their reactions to the iron- 

 haematoxyline and Hermann's stains, but resemble the true nucleoli of 

 Pentatoma in their reactions to the Ehrlich-Biondi and haematoxyline- 

 eosin stains. A study of these staining reactions accordingly öfters 

 no means of determining whether the nucleoli of Peripatus are related 

 to the chromatin nucleolus or to the true nucleolus of Insects; this 

 is not surprising, for the chemical nature of true nucleoli varies not 

 only in difterent cells, but frequently in different stages of the same 

 cell (cf. Montgomery, 1899). I think that in the spermatogonia and 

 spermatocytes of Peripatus occur only true nucleoli, and no structures 

 at all morphologically comparable to the chromatin nucleoli of Insects, 

 and for the following reasons: 1) There is only one kind of nucleoli 

 in Peripatus, and these seem to disappear in the prophases of division ; 

 further there does not occur in the 1st maturation division any special 

 chromosome which difi'ers in volume and form from the other chromo- 

 somes, whereas in the Insects in the 1st maturation monaster the 

 chromatin nucleolus is easily distinguished from the other chromosomes 

 either by its volume or by its form. 2) The nucleoli in Peripatus 

 differ in form and number, whereas the number of chromatin nucleoli, 

 at least so far as yet made known, appears to be constant to a cell. 

 3) When the nucleoli of Peripatus first appear in the synapsis they 

 are strictly peripheral in position, exactly as is the case with the 

 true nucleoli of Pentatoma. 4) In Peripatus there is present a peri- 

 nucleolar clear space, which is generally quite characteristic of true 

 nucleoli, but which I have not found around chromatin nucleoli. 5) In 

 Peripatus I have found no evidence of genetic connection of the 

 nucleoli with chromosomes. 



From all these considerations I think we may conclude that in 

 the spermatogonia and spermatocytes of Peripatus all the nucleolar 

 structures are true nucleoli, and that there are no bodies present com- 



