14 KDWAKI) T. UltoWNlv 



small opposite perrailial tentacles at an early stage of development, witli filaments 

 just appearing. Four interradial marginal liullis and eight adra<lial bulhs, smaller 

 than the interradial. 



Size. — Umbrella aliuut 1 mm. in width ami height. 



The presence of two long oppo.site perradial tentacles in the early stage indicates 

 that this Medu.sa begins its frce-awimming career with only two tentacles. The genus 

 Catablema is very closely related to the genus Tiara. It is known that Tvira pileata 

 is liberated from a hydroid belonging to the genus Perlt/oniinus, and that on liberation 

 the .Medusa has only two opposite perradial tentacles. It is very likely that ('ataldema 

 is also liberated from a PerigonimusAWic hydroid. The early stages of Catabteiiui 

 ireldoni were taken in January and June, and the adults during April and May. 



The ' Di.scovery ' collection contains nine specimens of this new species, which 

 can be easily distinguished from the others of the genus by the tentacles po.s.se.ssing 

 filaments with nematoeysts, and by the length of the lateral diverticula of the radial 

 canals. The specimens all came from under the ice in McMurdo Sound. There are 

 two early stages with four tentacles, and two intermediate stages with 9 and 12 

 tentacles. The others are adults with 16 tentacles. Only two specimens are in good 

 condition. 



The ' Southern Cro.ss ' collection possesses ten spei-imens, all of which arc in a 

 very bad and rotten condition. They are solely recogni.sable by the structure of the 

 tentacles and by the lateral diverticula of the radial canals. These specimens, 

 however, were useful, for some are larger in size and pos-sess more tentacles than those 

 in the other collection. They were all taken at Cape Adare, at the surface and near 

 the beach, on 10th May, 181)9. Temperature of the .sea, 27° F. 



As the stomach is large, its attachment to the roof of the uiiibrellar aivity is 

 strengthened by "mesenterie.s." These so-called mesenteries are formed by outgrowths 

 of the stomach alon.j a portion of the radial canals, and <on.sefiuently the canals lejive 

 the stomach not at the top, but laterally. In Cataltlmna u-ildoni, the outgrowths are 

 very short, extending just over the top of the umbrellar cavity, and unless specially 

 searched for are likely to be overlooked. Prof. Haeckel attached importance to the 

 presence or absence of mesenteries in his cla.sSification of the Tiaridju, and included 

 them in the character of the genera. They are not true mesenteries, such as 

 Pti/c/uii/a.'itria polar is posses.ses, but simply outgrowths of the stomach, and their 

 extension along the canals depends greatly upon the size and weight of the stomach. 



The gonads (Plate I., fig. 5) are arranged in eight straight, adradial, longitudinal 

 rows, which extend along the whole length of the stomach. Each row is composed 

 of many small transverse fold.s, which bear the generative cells. The arrangement 

 of the gonads in straight rows is only seen in those specimens which have the stomach 

 properly expanded. Two specimens have their stomachs contracted back, and the 

 gonads arc curved and thrown back against the top of the umbrellar cavity. 



In Catablema aimpaiiula, Haeckel, the diverticula of the radial canals are short 



