332 BULLETIN 82, UNITED STATES NATIONAL MUSEUM. 



large numbers. The branches arising from the axial cord appear never to possess 

 a sheath of any kind. 



In Tropiometra carinata (or picta) Hamann found that a very different con- 

 dition obtains. While in Antedon it is only with certain reservations that we 

 can speak of a connective tissue sheath about the nerve stems, Tropiometra 

 possesses a verv thick and dense neurilemma which is best developed about the 

 cirrus nerves and least developed on the periphery of the central organ. 



In a cross section of a cirrus nerve the larger part is occupied by the neuri- 

 lemma, the smaller by the nerve itself, from which alone the nerve branches are 

 seen to arise. 



The neurilemma is composed of two different portions; for the most part it is 

 made up of fibers and cells lying very close together in the groxmd substance. 

 This fibrous part is separated from the nerve fibrillse by a membranous sheath 

 which incloses the nbrillse and extends outward into the fibrous portion as a 

 network. The neurilemma is sometimes sharply differentiated from the ground 

 substance of the ossicles, and sometimes passes directly into it. 



About the axial cords of the arms the neurilemma forms only a thin sheath, 

 but it passes directly through these cords as a longitudinal dorsoventral partition. 



The nerve branches given off from the axial cords are inclosed in a neurilemma 

 for only a short distance, and their branches are quite unsheathed. 



The structure of the neurilemma about the axial cords is the same as that 

 of the neurilemma about the cirrus nerves, but the former is never so highly or 

 perfectly developed as the latter. 



It is possible that the extraordinary toughness of the species of this genus, as 

 contrasted with the brittleness of the species of the Antedonidse, and the conse- 

 quent relative infrequency of regeneration processes, may in part account for the 

 presence of a strongly developed neurilemma, or rather that the necessity for 

 frequent reparation of lost parts in Antedon and Ileliometra may have resulted 

 in the reduction of the neurilemma in those forms. The great reduction in size 

 of the chambered organ in Tropiometra as compared with Heliometra may pos- 

 sibly be correlated with the high state of development of the neurilemma in the 

 former. 



EPITHELIAL NERVES. 



The epithelial nerves are composed of a continuous layer of nerve fibrillse, 

 which, truly epithelial in position, run between and at right angles to processes 

 from certain of the columnar epithelial cells forming the floor of the ambulacral 

 grooves of the disk, arms, and pinnules, sending out branches to the tentacles and 

 converging with the ambulacra to the mouth, thence passing inward under the 

 epithelium of the esophagus and of the digestive tube and, rapidly diminishing 

 in importance, following the further course of the latter as a very slightly 

 developed fibrillar layer. 



No true circumoral ring is formed by this nervous system, for all the fibrillse 

 converging from the ambulacral grooves maintain their original relative direction 

 so that about the mouth a funnel-shaped nerve sheet is formed, composed of 



