ZOOLOGY AND BOTANY, MICBOSCOPY, ETC. 55 



nutrient yolk appears to be the homologue of the endoderm of other 

 animals ; the sole difference of importance is that the endoderm does 

 not here take any active part in the formation of the embryonic body, 

 for even the enteric canal is formed in Aphis by the ectoderm ; this 

 divergence from the ordinary arrangement is explained by the size of 

 the endodermal mass, but it is likely to modify our ideas as to the 

 germinal layers. The mesoderm is the homologue of the typical 

 mesoderm, and the ectoderm to that of other animals ; in insects the 

 blastoderm, together with the invaginated portion which forms the 

 germ-bands, and the inner embryonic membrane is to be regarded as 

 the ectoderm. 



There seems to be no doubt that the breaking up of the nutrient 

 yolk into yolk-spheres is nothing else than the end of the cleavage 

 process, that the yolk-spheres represent endodermal cells, and that 

 they give rise to the so-called wanderiug cells. It is doubtful, how- 

 ever, whether these cells do really wander. 



In Aphis, where the ova are endoblastic, the mesoderm is formed 

 by the cleavage along its whole surface of the germ-band ; later, it 

 divides into a median and two lateral bands, which become seg- 

 mentally arranged; there are wanting, however, the lateral multi- 

 laminate parts, which are to be found in ectoblastic ova. The 

 difference in the history of the mesoderm of ecto- and endoblastic 

 eggs is due to the different mode of formation of the germ-band ; in 

 the latter there is a proportionately delicate involution which forms 

 the great part of the band, and as the mesoderm cannot be formed by 

 involution it is formed by cleavage ; in ectoblastic eggs the mesoderm 

 is ordinarily formed before the band, which is of greater width. The 

 heart of Aphis and its aorta are formed from a cord of mesodermal 

 cells, which is at first solid, increases by cell-multiplication, and 

 when it is hollowed out probably gives rise to the blood-corpuscles. 

 The author's observations agree in many points with those made by 

 Bobretzky on the heart of Crustaceans. The coelom of insects is 

 primary, for it arises from the cleavage-ca7ity. In Aphis, and 

 probably in all insects, the intestine is formed from an anterior and 

 a posterior ectodermal invagination, and there is no median arch- 

 enteron.* These invaginations are, from the first, invested by 

 mesoderm, which later on forms the muscles. 



The tracheae begin to be developed at a comparatively late period, 

 and this seems to be due to the small size of the germ-band, which 

 does not overlap the rudiments of the appendages in Aphis, or in 

 other insects with endoblastic ova ; the early appearance of the 

 tracheae in ectoblastic eggs is regarded as a secondary phenomenon. 

 The nervous system, again, presents differences in endoblastic and 

 ectoblastic eggs. 



We may recognize three periods of development ; in the first we 

 have the phenomena which precede the formation of the organs — 

 cleavage, formation of germinal layers, appearance of the germ-band 



* We may point out that Witlaczil is in error in supposing that Prof. Balfour 

 was the author of the words stomodseum and prootodseum ; we owe these two 

 useful terms to Prof. Kay Lankester. 



