THE EMBRYOLOGY OF THE HONEY BEE 153 



an ordinary cup-like invagination greatly drawn out and flattened. 

 At stages prior to hatching the long axis of the invagination is 

 oblique to the long axis of the embryo, so that both coronal and 

 transverse (Fig. 58A-C) sections intersect the long axis of the 

 invagination. At hatching the dorsal flexure of the embryo is 

 exchanged for a ventral flexure, and the head becomes turned 

 ventrad, so that the long axis of the invagination is brought to a 

 position approximately at right angles to the long axis of the 

 young larva (Fig. 39). 



The accounts of the formation of the optic ganglion and optic 

 plates of insects differ to a considerable extent. In Mantis (Vial- 

 lanes 1891) the optic ganglion is produced by "Cellules ganglio- 

 genes" or neuroblasts like the remainder of the brain, the over- 

 lying hypodermis splitting off from the optic ganglion to 

 constitute the optic plate. In Xiphidium (Wheeler 1893) the 

 optic ganglion and optic plate are formed by a simple separation 

 of two cell layers, as is Mantis, but the inner layer is single and 

 contains no neuroblasts. In Forficula (Heymons 1895) the pro- 

 cess is much the same, but Heymons reports having observed a 

 few neuroblasts in the layer forming the optic ganglion. In none 

 of these forms is there any invagination of the ectoderm forming 

 the optic plate. There is present however, an ectodermal invagin- 

 ation or proliferation just posterior to the embryonic optic lobe 

 (Fig. 38, Igl), but this is outside of the limits of the latter and 

 situated between it and the second protocerebral lobe, and is 

 moreover purely hypodermal in character, so that as will appear 

 later there seems to be no good ground for comparing this in- 

 vagination, the intergangionic thickening of Wheeler (1893), with 

 the invaginations of the optic lobe described by Patten (1889), 

 as Viallanes (1891) and Wheeler (1893) have done. In the re- 

 maining accounts, which relate exclusively to the orders Coleop- 

 tera and Hymenoptera a conspicuous ectodermal invagination is 

 concerned in the development of the optic lobe. Patten (1887) 

 was the first to observe this in the case of the wasp Vespa. Here 

 the rudiments of the eye are formed somewhat as in the honey 

 bee, the optic plate lying at first laterad of the rudiment of the 

 optic plate, and later extending mesiad over it. The separation 

 of the optic plate from the optic lobe however takes place early, 



