l6 SMITHSONIAN MISCELLANEOUS COLLECTIONS VOL. ']'] 



facts either of the development or of the structure of the parts 

 concerned. The median gangHa of the stomatogastric system, more- 

 over, are so variable in number that the scheme given above can 

 scarcely be regarded as representing a primitive arrangement. The 

 frontal ganglion is the only one of constant occurrence ; one or 

 both of the other median ganglia may be lacking. The lateral 

 ganglia must in any case be regarded as of secondary origin. They 

 are apparently always connected with the median system, but each 

 has also one or two connectives with the back of the brain. In 

 caterpillars the lateral ganglia are entirely separated from the walls 

 of the alimentary canal, and their principal nerves go to muscles 

 of the anterior dorsal and lateral parts of the head. 



According to the accounts of Zawarzin (igi6) and Orlov (1924), 

 the nerves of the stomatogastric system contain both motor fibers 

 and sensory fibers, the former having their cytons in the ganglia, 

 the latter arising from cells distributed over the innervated parts 

 of the alimentary canal and ending in fine terminal arborizations 

 within the ganglia. 



THE ORIGIN OF THE SENSORY NERVES 



Since no sensory cytons have been found in the central gan- 

 glia of insects, or anywhere associated with them, students of the 

 central nervous system follow vom Rath (1896) in insisting that 

 the generative cells of the sensory nerves must be found in the 

 periphery, presumably in the ectodermal tissue of the body wall 

 and the alimentary canal. 



When the sensory fibers are traced outward from the central 

 ganglia, each is found to end in a cell. Some of the cells are 

 bipolar, others are multipolar. The distal process of some of the 

 bipolar cells goes direct to a sense organ of the ectoderm, either 

 one in the hypodermis, or one in the epithelium of the alimentary 

 canal. Bipolar cells of this kind, connected with specific sense 

 organs, constitute the sensory cells of Type I of Zawarzin (1912 a). 

 The other cells, which may be either bipolar or multipolar, but 

 which are characteristically multipolar, give off terminal processes 

 that end in fine branches on the inner surface of the hypodermis, 

 perhaps also on the tendons of the skeletal muscles, and on the walls 

 and muscles of the alimentary canal. Cells of this kind constitute 

 the sensory cells of Type II of Zawarzin. The cells of both types 

 lie either within or just beneath the ectodermal tissues they inner- 

 vate, and they are the only cells yet found in the courses of the 

 sensory nerves. This has given rise to the idea that they are the 



