20 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. 'J'J 



vulgaris (fig. 8B). Zawarzin (1912) says also that there are branch- 

 ing nerve fibrils in some places beneath the membranous parts of the 

 cuticula in dragonfly larvae (fig. 8 A), and Monti reports the presence 

 of a rich arborization of nerve fibers arising from multipolar cells on 

 the inner surface of the hypodermis in adult Orthoptera. Several 

 writers have described a nerve net beneath the body v^^all of other 

 arthropods; Bethe (1896), for example, found it in the freshwater 

 crayfish, Nemec (1896) in land isopods, where, he says, the nerve 

 endings lie between the hypodermal cells, and Holmgren (1896) re- 

 ports a plexus similar to that of insects in various groups of Crusta- 

 cea. 



The writers mentioned above all assert that the cells of the subhy- 

 podermal net are nervous elements, but Duboscq (1897) claims that 

 similar cells present in Forficulida; are connective tissue cells, and 

 he discredits the idea that the cells are in any case nerve cells. All 

 studies of the subhypodermal network and its cells have been made 

 by the methylenblue method of staining nerve tissues. 



The most concise account of the subhypodermal innervation of an 

 insect is that by Zawarzin (1912 a) made on the larva of Melolontha 

 vulgaris. Beneath the skin of the larva, Zawarzin says, there is a 

 finely-branching system of nerve fibrils forming a network of large 

 and small meshes over all parts of the body (fig. 8B), including 

 the appendages, but particularly developed on the middle of the 

 back. The nerves (Nv) that break up to form the net proceed from 

 small, irregular cells of Type II (Cyll), some bipolar, other multi- 

 polar. The distal processes of the cells branch dichotomously into the 

 fibrils of the larger meshes, and these ramify to form the threads of 

 the finer meshes. None of the fibrils, Zawarzin says, unite with one 

 another, though they often appear to do so when they lie close to- 

 gether. Monti noted the same in studying the subhypodermal nerve 

 net of Cerambycid larvae. The fibrils of the larger meshes in Melolon- 

 tha are relatively smooth, but the finer branches are characteristically 

 varicose, presenting numerous small swellings along their courses. 

 All investigators have noted this varicosity of the end branches of 

 the subhypodermal nerves of insects, and the same feature is described 

 for the branches and fibrils of the sensory nerves in the human epider- 

 mis. The actual endings of the fibrils in insects have not been defi- 

 nitely observed, but the terminal branches appear to end free on the 

 basement membrane. 



The character of the fibrils in the subhypodermal nerve net appar- 

 ently leaves no doubt that they are the terminals of sensory nerves, 

 since the fibrils of the sensory roots in the central ganglia have the 



