1!28 DR. E. KLEIN. 



bordered by a definite, though delicate, membrane. There 

 are, however, present in the parablast also true cells, i. e. 

 granular corpuscles (more transparent than the parablast 

 itself), containing two, three, or more bodies, which, either all 

 or only some, possess a complete similarity with the nuclei of 

 the neighbouring p^rts of the parablast. That masses of 

 substance of the parablast separate themselves from the rest 

 and collect round the nuclei so as to form " nucleated cells " 

 is, I think, the simplest way of explaining the facts just 

 stated. The question at present of greatest importance is, 

 how do the first nuclei of the parablast develop ? The cir- 

 cumstance that nuclei are not perceptible in the parablast in 

 the earlier stages of segmentation, as mentioned on a former 

 page, although not absolutely proving, renders it at least 

 very probable that they originate in situ, as development in 

 general proceeds. Besides, there is one definite fact I have 

 been able to ascertain in somewhat later periods {e.g. in 

 preparations like the one represented in fig. 5), which is apt 

 to throw some light on the origin of these nuclei. It is this : 

 searching carefully through the substance of the parablast 

 with a moderately high power (Hartnack's No. 8), we detect 

 numerous isolated, small, transparent bodies very faintly 

 outlined, so as to be rendered just perceptible ; between these 

 and distinct nuclei all intermediate forms may he met with as 

 regards getieral aspect, outline, and size. This obviously 

 means new formation of nuclei. It therefore stands to 

 reason to assume that, inasmuch as at a period when nuclei 

 may be seen to multiply by division the formation of nuclei 

 de novo, as it were, still takes place in the parablast, the 

 first nuclei of the parablast have also originated hi the same 

 maimer, i. e. de novo. 



From the fact that some of the nuclei of the parablast 

 possess elongated or knob-like processes of different sizes, 

 which in some instances are more or less constricted off" from 

 the main mass of the nucleus, it may be said that the multi- 

 plication of nuclei takes place also by gemmation or budding. 



The appearances of multiplication described by Balfour of 

 the " nuclei of the yolk " in elasmobranch fishes (1. c, p. 4 ; 

 compare also fig. 1 of his paper), as quoted on a former 

 pace, are not identical with those of the parablast of our 

 case ; whereas the appearances described and figured by 

 Oellacher^o (figs. 30—36) of the nuclei of the archiblast 

 during early segmentation more closely correspond with 

 those of the nuclei of our parablast. 



(b) On a previous occasion we mentioned that on account 

 of the intimate connection, at certain points, between the 



