I. INTRODUCTION. 



The recent and extensive investigations of the intimate structure of nervous 

 cytoplasm and of the changes occurring in it, dating from the use of the aniline dyes 

 for staining purposes by Nissl and others, have yielded vast and profoundly important 

 results. The work has been done almost entirely on man and the higher vertebrates. 

 The present research was undertaken to find out if similar conditions exist in inverte- 

 brates. 



In the higher animals the motor cells have been the easiest ones to study in this 

 connection. The cockroach, therefore, on account of its active habits and abundance 

 was available invertebrate material for such work. Its nervous system, following 

 the arthropod type, consists of supra- and sub-cesophageal ganglia united by circum- 

 cesophageal connectives, and a series of single segmental ganglia lying in a median 

 ventral position and connected, each with the next, by a double connective. There 

 are three thoracic ganglia, larger than the abdominal ones, and presumably control- 

 ling, among other organs, the three pairs of legs. These three ganglia were used 

 indiscriminately in studying the cells. These ganglia, like the connectives, are 

 enveloped in a sheath of areolar tissue; they are dorsoventrally flattened and show 

 a tendency to division into symmetrical lateral halves. Their central and major 

 portion is formed of interlacing fibres, while on their peripheries, especially on the ven- 

 tral surface near the anterior and posterior ends, he the nerve-cells embedded in loose 

 areolar tissue. The cells vary greatly in size, their longest diameters being from 

 10 to 80 micra; they are oval and regularly unipolar, only one bipolar cell having 

 been observed. No dendritic processes were seen, but fine ones may exist, as no 

 impregnation preparations were studied. 



II. CELLS FIXED IN VOM RATH'S OSMIC FLUID. 



i. Technique. — A living roach was decapitated, opened along the back, and 

 the ganglia dissected out under water, the manipulation taking eight minutes. The 

 animal moved its legs for two or three minutes after decapitation. (The same method 

 was used in getting out all the subsequent material, the only modifications being 



341 



