Miscellaneous. 75 



In Meloe, the species of beetle studied, probably a corresponding 

 stage occurs in which all the cells are at the surface, though there 

 are no yolk-pyramids ; consequently, in the grasshopper and in 

 Meloe the yolk-cells probably arise by delamination from the cells 

 investing the yolk. 



The erabrj'o of Tlnjridojjteryx and of other insects studied arises 

 as a thickening on the surface of the egg not unlike the primitive 

 cumulus of spiders. 



The amniotic folds arise as folds of blastoderm on all sides of 

 the embryo, and finally meet and unite over the median line of the 

 ventral plate ; consequently the embryo (described as the ventral 

 plate at this stage) comes to lie in the yolk covered by the inner 

 amniotic fold or true amnion, while the outer fold or serosa remains 

 continuous with the blastoderm. The embryonic membranes of 

 Mantis and Meloe arise in a quite similar manner. Brandt has de- 

 scribed a different mode of origin for the embryonic membranes of 

 the Neuroptcra and Hemiptera. 



After the formation of the membranes in Thyridopteryx, but 

 synchronously with the same in Mantis and Meloe, an ingrowth 

 occurs in the middle line of the embryo, which is partly a dela- 

 mination and partly an invagination. By this ingrowth is formed 

 the inner germ-layer, which in Thyridojyterycc certainly corresponds 

 to both mesoderm and endoderm. The yolk-cells do not appear to 

 take any part in the formation of the endoderm in Thyndopteryx. 

 TichomirofF, from his studies in the Lepidoptcra, comes to a similar 

 conclusion in regard to the yolk-cells. 



The yolk-cells of the grasshopper also appear to take no part in 

 the formation of the endoderm. 



The amnion in Thyridopteryx grows dorsally more rapidly than 

 the body-walls and its opposite folds unite dorsally before the body- 

 walls can grow together. Consequently the amnion in this insect 

 forms part of the dorsal surface of the body, while for a time the 

 entire embryo is enclosed as in a sack by the outer fold of the 

 true amnion, which does not take part in the closure of the dorsal 

 surface. 



Xo dorsal organ corresponding to that described by Brandt for 

 the Neuroptera was observed in Thyridopteryx or in the other 

 insects studied. The amnion of the grasshopper does not apparently 

 form any considerable portion of the dorsal wall of the bod}'. 



The nervous system arises in all insects studied as two ectodermio 

 strings lying on each side of the blastopore, as the median line where 

 the inner layer arises may be called. It subsequently divides into 

 a number of ganglia corresponding to the somites of the body. The 

 supraoBsophageal ganglion, as good longitudinal sections of the 

 TJiyridopter yx-emhryo show, consists of two portions — a posterior 

 portion which innervates the paired labium, and the anterior portion 

 which supplies the antennae with nerves. The circumoesophageal 

 commissure is formed by a portion of the posterior division of the 

 supraoesophageal ganglion and a portion of the mandibular division 

 of the Buboesophageal ganglion. The supraoesophageal ganglion of 



