34 T. KOMAI: STUDIES ON TWO ABERRANT CTENOPHORES 



chromatin substance on the surface — this is probably the spermatid stage. 

 The spermatozoon (s/^3, fig. 3) is filiform, with a very slender head 

 tapering anteriorly a little and measuring some 7 /j. in length. The head 

 of the spermatozoon of Bero'e, as shown in Chun's work ('80, PI. 16, fig. 

 49), is nearly pyriform, it differs thus rather strikingly from that of the 

 spermatozoon of Coeloplana. 



Besides the sexual cells of various developmental stages described 

 above, there are found in the testicular cavity cells that are apparently 

 non-sexual in character (PL 5, figs. 2, 4, 8; n.-s. c). They are found 

 among sexual cells and may be observed with utmost clearness when 

 imbedded among the tails of spermatozoa. They resemble in appearance 

 the spermatogonia fairly well, but may be distinguished by the greater 

 size (3-5 yL in diameter) and clearer appearance of the nucleus and also 

 by the c}'toplasm staining more strongly with eosin. . Owing to the 

 character mentioned last, the cells often stand out very conspicuousl\- 

 among the various elements in the testis. Of the three Coeloplana 

 .species, C. willeyi and C niitsukurii seem to be provided with more 

 numerous cells of this kind than C. bocki. The nature of those cells 

 remains problematical to me, besides that, they perhaps perform some 

 nutritive function for sexual cells. Such cells have not been described 

 from any ctenophore so far. 



The testis is covered by a membrane (PI. 5, figs, i, 2, 4, 8; te. in) 

 which limits it from the surrounding parenchyme tissue. The membrane 

 varies in thickness a great deal and often becomes so thin as to show 

 an almost structureless appearance, although it never fails to reveal its 

 nuclei to closer examination. Apicall}^ it continues into a tubular 

 structure which opens to the dorsal surface of the body (figs. I, 2, 8; 

 V. d). This structure is made up of fairly tall ciliated cells arranged 

 around a rather spacious lumen. In the cross-section of the tube, one 

 can count about twenty such cells (fig. 8). The lumen is filled with a 

 sperm mass frequcntl)-, and in one case have I observed a strand-like 

 mass of sperms to lie just at the opening of the tube, partially outside of 

 the opening and partial 1\- in the lumen of the tube. These observations 

 seem to indicate that, the tube functions actually as the sperm-duct. At 

 the opening of the tube, the e[)idermis often shows a slight depression 

 (fig. 8). 



In the testes of the earliest stage of development, the duct is 



represented by a short finger-shaped process from the apical part of the 



• testis which terminates some distance from the dorsal surface of the body. 



