GENERAL CONSIDERATIONS. 145 



layers of entodenn-cells, which lie below the germ-band, and, as the 

 cells increase in number by division, spread out gradually over the 

 surface of the yolk until they completely surround it. The entoderm 

 has here, as explained on p. 136, divided into two parts, one plastic 

 and the other transitory. We may further recall that in Type II. 

 also, not all the wandering cells scattered through the yolk came 

 to the surface (p. 132) to take part in the formation of the mid-gut 

 epithelium, but that some of them remained in the yolk and finally 

 disintegrated. Such cells evidently correspond to the vitellophags 

 in this type. The actual key to the processes here described is to 

 be found by a close examination of the method of development 

 already described for Astacus. ^Ve have shown (p. 129) that the 

 cells of the entoderm-vesicle in Astacus do not participate alike in 

 the absorption of the food-yolk. Those of the dorsal half are most 

 concerned in this process, while those of the ventral half are less 

 aflfected by the filtration of the yolk. It is from the latter cells, 

 however, that the formation of the definite mid-gut at first proceeds. 

 We find first, near the blind end of the proctodaeum, a plate of 

 entoderm-cells (Fig. 63 B, ep), which already shows the distinctive 

 characters of the mid-gut epithelium, and has a certain tendency 

 to overgrow the other unmodified parts of the entoderm. A 

 precisely similar entoderm -plate is also developed in Type II. 

 (Fig. 64 C, ep), so that there also some of the entoderm-cells 

 scattered in the yolk show greater plastic capacity than the rest. 

 We thus find here a beginning of a division of labour which is 

 completely developed in Type III. {cf. p. 136). To Type III. 

 belong the Mijsidae, the Arthrostraca, and the Cumacea (V). This 

 type is found also in other divisions of the Arthropoda, modified 

 in various ways ; it is met with, for instance, in the Scorpiones and 

 in the Insecta. 



Mesoderm. Only in the small egg of CetocMhis does the meso- 

 derm arise from paired primitive mesoderm-cells. In most Crustacea 

 the rudiment is from the first multicellular. Among the many 

 accounts of the origin and position of the mesoderm in the various 

 groups of the Crustacea, its first appearance (in the Decapoda) at 

 the anterior part of the lip of the blastopore may be regarded as 

 a comparatively primitive process, from which it is possible to 

 deduce the process described for Ligia (p. 136), and, through this 

 form, perhaps, that in many other Crustacea. 



It is a striking fact that there is very little tendency in the cells 

 of the mesoderm from the very first to form regular layers. JNIere 



L 



