286 BULLETIN 100, UNITED STATES NATIONAL MUSEUM. 



Canals. — The margins of the radial canals are lobed in their mid- 

 region, but their outer ends and the circular canal are smooth. 



Color. — In the preserved specimens the tentacles, manubrium, and 

 gonads are pale yellow. 



Family BYTHOTIARIDAE Maas, 1905. Sens. em. Bigelow 

 (1909a, 1918), Maas (1910). 



Tribe Bythotiaridi (part) +Tribe Calycopsidi Mayer, 1910. " Gruppe " 

 Calycopsiden Hartlaub, 1913. 



Hartlaub (1913, p. 349) substitutes the name " Calycopsiden " for 

 " Bythotiaridae," both because the former is derived from the oldest 

 and largest genus, and because he finds the latter "unbequem." But 

 neither of these reasons seems to me to justify abandoning a gen- 

 erally accepted family name. 



Maas (1910) has recently given us so satisfactory a discussion of 

 this family that I need add little here beyond the statement that I 

 am still in thorough accord with his union of Bythotiara, Hetero- 

 tiara, Sihogita, and Calycopsis in a single family, distinct on the one 

 hand from the Williadae, on the other from the Pandeidae. Mayer 

 (1910), on the contrary, has referred Bythotiara and Sibogita to the 

 Williadae, Heterotiara to the Pandeidae (" Tiarinae "). Vanhoffen 

 (1911), who has examined more specimens of the three genera than 

 any other student, has classed them all among the Pandeidae 

 ("Tiaridae"), but without any discussion of their relationships. 

 And Hartlaub (1913) in his recent revision of the Pandeidae 

 [" Tiaridae "] classes them as " Gruppe Calycopsiden " of that 

 family. But although, as I have previously pointed out (1909a, p. 

 213), "the closest relationship of the Bythotiaridae is undoubtedly 

 with the Tiaridae [Pandeidae]," they are so easily distinguishable 

 from the members of that family by the greater development of the 

 manubrium, which, as in most Leptomedusae, is distinguishable into 

 basal, gastric, and oral regions; by the structure of the gonads, which 

 are permanently interradial instead of having this primitive loca- 

 tion masked with growth as is the case in most Pandeidae; by the 

 structure of the bell margin and by the fact that the tentacles have 

 no distinct bulbs; and they are so uniform, among themselves in 

 these respects, that I still believe they are best grouped in a separate 

 family. But here, as so often among Medusae, it is perhaps im- 

 possible to draw a hard and fast line; for, as Hartlaub (1913) points 

 out, the genus Meator, recently described by me from the northwest 

 Pacific (1913) has some characters in common with the Pandeids, 

 some with the Bythotiarids. 



A feature of special importance emphasized by Maas is the ab- 

 sence of any swelling at the base of the tentacles. 



