NKW Sl'l':('ll':s (•!•' CK'AWI.INd MKhlSA ri.'oM Al'STK'A I.IA KIMUiJS. 101 



aiul gland cells, but uo.iieniatcK'ysts siicli as (j ilclnisl louiid in tin- wall of 

 the .stomach of C. cupense. The gland cells are veiy widely distrihutt-d 

 thi'onglioiit tlie endoderni. 



Tlie cavity of the stoniacli is picKlnced into eight angular enil)ayinents 

 forming tlie stomach diverticula, which are narrow, acute-angled pouches 

 lying between the longitudinal endodennal ridges. These diverticula give 

 a very charactei-istic star-like appearance to transvei-se section.s through 

 the region of the stomach. Fi'oin the apex of each of tlie stomach diver- 

 ticula is given off a short radial canal, which passes directly to the 

 circular canal. 



The ai'raugement of the diverticula in ( '. Imxin-Ui agrees exactly with 

 Vanhoffen's Hguie of a transveise section through a male specimen of the 

 Kerguelen species. In C. cdpeiixe, on tlie other hand, the stomach diver- 

 ticula are always six in number and take the form of wide pouches, which 

 are more oi' less rectangular when fully developed, although in younger 

 specimens they may be small and acute. 



The stomach diverticula in ('. Iiustrelll are eight in number in 

 all except two young individuals, in which there are six and seven 

 respectively. 



Fitdidl CKinils. — The radial canals are usually eight in number ; of two 

 3'oung individuals, liowever, one had only six, while the other, which was 

 more mature, possessed seven. The ladial canals are very short, and being 

 uupigmented or nearly so, can only be made out with certainty in sections. 

 An examination of the living animal as well as preserved specimens 

 cleared in cedarwood oil did not reveal the true arrangement of these 

 canals. Owing to the stomach diverticula reaching almost to the circular 

 canal, there is, in the whole specimens, an appearance of pigmented radial 

 canals ; in sections, howevei', the apices of these diverticula are seen to 

 communicate with the circular canal by very short radial canals, the 

 endodermal lining of which is devoid of pigment-granules. 



The number of radial canals present in the several species of 

 ( 'in'doiienia varies very considerabi}'. In the case of ( '. vdlleutliii, Browne 

 states that " according to Mr. Vallentin, the Falkland species has four 

 radial canals," and that ('. charcotl (Bedot) is distinguished from all the 

 other crawling Medusae of tlie Southern Hemisphere '' by the radial canals 

 having slender lateral branches with a tendency towards anastomosis." 

 The latter statement by Browne is based presumably on his examination 

 of specimens of Wainlelin clntrrntl received from Professor Bedot. 



The radial canals of ('. li(idij,<o)ii are variable in number. Browne 

 found that out of si.v specimens examined, three had eiglit radial 

 canals and the remaining three had six, ten and eleven canals 

 respectively. Vanhoffen does not state the number of radial canals 

 present in the Kerguelen specimens, but an inspection of his figure of a 

 transverse section througli a male shows eight radial canals. He describes, 

 however, the endoderni of the stomacdi as forming six, eight or ten folds 

 which come into communication with the circnlai- canal by short radial 

 canals. This ari-angement of the stoniacli diverticula w^ould seem to 

 indicate that the radial canals are ahso variable in number, and are 



