THE EMBRYOLOGY OF THE HONEY BEE 141 



time as the delamination of the lateral cords, its interganglionic 

 portions remain in connection with it, after the nerve fibres have 

 become evident in the lateral cords." 



Escherisch (1902, 19023.) has given a very complete and in- 

 teresting account of the development of the median cord in 

 Musca, in which the conditions recall those found in Lepisma. In 

 J\Iusca a continuous median cord is separated from the ectoderm. 

 Within the limits of the ganglia the median cord contributes their 

 median portions, as in other insects ; in the interganglionic re- 

 gions it presents swellings of considerable size, one being situated 

 directly caudad of each ganglion. From each of these swellings, 

 near its posterior end, a pair of lateral processes are given off 

 which extend to the neighborhood of the stigmata. These pro- 

 cesses are regarded as nerves. Escherisch points out the close 

 resemblance which the median cord bears to the unpaired median 

 nerve described by several investigators of the anatomy of insects. 

 This discovery, as well as that of Heymons (1897), shows that in 

 some insects the median cord may form a more or less continuous 

 median nerve. The presence of such a nerve in Lepisma suggests 

 that its occurrence represents a primitive condition. 



Hirschler (1909) in the chrysomelid beetle Donacia, has in part 

 confirmed Graber's statements regarding Melolontha, since he 

 finds that the median cord is at first completely severed, through- 

 out its extent, from the hypodermis, but that afterward the inter- 

 segmental sections are added to the ganglia, forming in each the 

 posterior median gangliomere. In their fate these interganglionic 

 sections correspond to the median cord neuroblasts and their 

 products in the Dermaptera and Orthoptera. 



In the development of the lateral cords it is evident that the 

 honey bee conforms to the general rule and that such differences 

 as it presents are of relatively minor importance. In the origin 

 and fate of the median cord it conforms to Grassi's account and 

 is also similar to many of the Coleoptera, as, for example, Hydro- 

 philus (Graber 1890), in so far as the median cord is not de- 

 veloped in the interganglionic regions, but remains here united 

 with the hypodermis. These interganglionic spaces nevertheless 

 are, up to the time of hatching, of very slight extent in an antero- 

 posterior direction, as figure 5oE shows ; moreover the anterior 



