Evihryology of the Phalangiiclse.. 399 



that blastodermic momhranc wliicli is produced from the 

 blastoderm in many Crustacea (figs. 7 and 11). 



4. The germinal disk arises at one pole of the ovum by 

 iTiultij)lication of the ectoderm cells. The newly formed 

 lower layer of the primitive streak represents the mesoderm, 

 since the endoderm is differentiated from the beginning. 

 Among the cells of the lower layer a grou]) is separated off 

 from the commencement, the cells of Avhich are distinguished 

 by their size and peculiar a|)pearance. The separation of 

 this group of cells even precedes the formation of the primi- 

 tive streak ; as early as the time when the ectoderm clothes 

 the ovum with a cellular layer this group of cells already 

 projects as a little cluster into the interior of the ovum (figs. 9, 

 10, 11). This cluster lies, as is subsequently to be seen, in 

 the posterior portion, although not quire at the end, of the 

 ventral streak, and consequently forms a local thickening of 

 the ectoderm, which arises almost simultaneously with the 

 mesoderm, and afterwards furnishes the germ-cells. 



5. The nuclei of the large endoderm cells frequently 

 suffered from the effects of the reagents, and then appeared 

 to be destitute of a membrane (fig. 8) ; but they were well 

 fixed by means of Flemming's fluid, and presented the 

 appearance shown in figs. 7, 9, 11, 12, and 13. The nuclei, 

 which are figured in lienking's paper mentioned above, also 

 seem to me (at least in some cases) to have suffered from the 

 fixative fluids, and therefoi'c to exhibit no membrane and no 

 sharp outlines. That which, for instance, he considers to be 

 several nuclei in one cell {vide his fig. 37), I am inclined to 

 regard as being nucleoli of a large nucleus, the membrane of 

 which is destroyed. At the time of the formation of the 

 mesoderm the nuclei of the endoderm become considerably 

 larger, so that in comparison with the cells of the germinal 

 disk they appear quite gigantic. They possess a sharp con- 

 tour and are very poor in chromatin ; almost the whole of 

 the colourable substance of the nucleus is concentrated in a 

 nucleolus, which is very glistening and takes a deep stain. 

 We often meet with figures which seem to point to amitotic 

 nuclear division (fig. 13) ; it appears that this nuclear division 

 is also followed by division of the cell (fig. 12). At any rate 

 the endoderm cells never become multinucleai*, and even cells 

 with two nuclei are rare. I succeeded in determining a 

 similar characteristic nuclear structure in the endoderm (yolk- 

 cells) of the Araneid^e also, in the earlier stages of their deve- 

 lopment; this had not previously been described by any 

 author {Tegenaria, figs. 14 and 15). In Araneida3 and 

 Plialangiidse there consequently occurs a fragmentation of the 



