THE DEVELOPMENT OE LEPIDOSIEEN FARADOXA. 25 



Size of Nuclei during Early Stages of Develop- 

 ment of Lepidosiren. — Owing to the small scale of the 

 figures it is not possible to indicate the relative sizes of the 

 nuclei in different parts of the egg. These bear, as one . 

 might expect, a rough relationship to the volume of the cell 

 territories over which they preside ; e.g. in two eggs of 

 Stage 16 the nuclei of the ectoderm averaged "016 mm. and 

 •014 mm. in diameter, those of the mesoderm '018 mm. and 

 •016 mm. J and those of the large yolk-cells "022 mm. and 

 •021 mm. Again, in an egg of Stage 13 the nuclei in the 

 region of the dorsal lip of the blastopore measured "015 mm., 

 and those ofc' the large yolk-cells •019 mm. 



The measurements ai-e in all cases the average of ten 

 measurements of whole nuclei as seen in thick sections. 



General Eb;MARKS. 



Segmentation. — In studying the segmentation of Lepi- 

 dosiren I have been much struck by the readiness with which 

 all trace of the division planes may be destroyed in the parts 

 of the egg filled with large yolk-granules. The two com- 

 monest causes of this are, firstly, the use of a fixing agent of 

 inferior penetrating power, the blastomeres running together 

 into a continuous mass very soon after death if the fixing 

 agent has not reached them ; and secondly, the use of two 

 thin sections. In cutting a section it would appear that the 

 yolk-granules become very slightly displaced as they strike 

 the edge of the knife, and if the section is very thin this is 

 enough to completely obliterate the division planes. During 

 segmentation in Lepidosiren thick sections will show an egg 

 to be completely divided up into blastomeres, while in thinner 

 sections the whole lower portion with coarsely-grained yolk 

 seems to form a quite continuous unsegmented mass. The 

 mass of uncleaved yolk figured by Semon in the middle of the 

 Ceratodus egg, and upon which he bases the statement that 

 this egg in its early stages of segmentation occupies a place 

 intermediate between the telolecithal and centrolecithal types, 

 may, I think, quite possibly be an artefact of this nature, due 



