328 BULLETIN 82, UKITED STATES NATIOIS^AL MUSEUM. 



bases and are produced into extremely fine fibers which are in connection with 

 the nerve fibrillae. These specialized groups of epithelial cells are found on the 

 dorsal and lateral surfaces of the brachials and pinnulars. Hamann found them 

 most easily demonstrated in Heliometra glacialis. 



The two ventral branches ramify in exactly the same way. One of them, 

 alternately the right and the left, enters the j^innule, in which its course and its sub- 

 division exactly duplicate that of the axial cord itself within the arm. After 

 very numerous divisions these ventral branches resolve themselves into fine twigs, 

 some of which reach the body epithelium, in the ambulacral lappets and on the 

 ventral portion of the arm generally, while others, radiating out brush-like into 

 single fibrilhv, enter the ventral interbrachial muscles. 



In Heliometra glacialis Hamann found in the portion of the axial cord 

 traversing the muscular articulations between each pair of brachials ganglia 

 composed of a niunber of unipolar cells, each with a globular body 0.01 mm. by 

 0.02 mm. in size containing a clear vesicular nucleus with a nuclear body, which 

 taper into a strong process penetrating the nerve mass in which it radiates out 

 like a brush. The ganglia have not been described in Antedon. 



In those portions of the calcareous ground substance which border immediately 

 upon the ligament masses of the muscular articulations, the articulations between 

 the cirrus segments, the syzygies and the synarthries lie numbers of the so-called 

 pear-shaped cells which give off at the small end often several processes pentrating 

 far into the fiber mass. The cytoplasm, including that of the finest divisions 

 of the processes, is very finely granulated. Reichensperger suggests that the 

 occurrence and strong staining of these granules recalls Nissle's bodies. 



These pear-shaped cells were first described as ganglia by Perrier. His suc- 

 cessors either doubted their nervous nature or failed to ascribe to them any definite 

 function until very recently Reichensperger, studying the question anew, was able 

 to confirm Perrier's interpretation of them. 



According to Reichensperger the only difference between these cells and the 

 ganglion cells of the axial cord lies in their sti"onger microchemical reaction, and 

 perhaps also in their more branched processes. The latter are also more or less 

 isolated, while these are massed, lying generally in pairs on the bows of the 

 dorsal ligaments. 



In Antedon thej'' recede somewhat from these bows, and the nerve twigs 

 therefore lie finely divided below them instead of above. 



The processes of the pear-shaped cells are not united with cells, but run 

 between the ligament masses and surround single bundles. The connection 

 between these pear-shaped cells and the axial nerve cord of the arms was worked 

 out in detail by Reichensperger. 



Just within the anterior and posterior border of each brachial there run out 

 from the axial cord nerve branches comparable to the nerve branches in the middle 

 of each brachial, but very much finer. These divide almost immediately and, 

 ramifying in a plane at right angles to the axial cord, resolve themselves into 

 extremely fine fibrillar bundles, which unite with the terminal layer of pear- 

 shaped cells. In a normal adult but few ganglion cells are found at the base 



