42 University of California Publications. [zoology. 



pieces of the stem in the following manner: The proximal 

 tentacles, averaging about twenty-four in the adult, arose as 

 buds, in two series. The tentacles of the first series appeared 

 simultaneously to the number of seven to fourteen, the number 

 being conditioned apparently by the diameter of the stem and the 

 number of canals in it; a tentacle arose on each canal. The 

 tentacles of the second series appeared singly, between the tenta- 

 cles of the first series, some time after the latter. 



The distal tentacles are filiform in the adult, and are scattered 

 and more numerous than the proximal ones; in small regenerating 

 pieces they are somewhat capitate at first and arranged more or 

 less regularly in whorls of four (quartettes), each whorl being 

 proximal and alternating with the one immediately preceding it. 

 The tentacles of the several whorls did not develop simultane- 

 ously. They may appear one at a time, in no order that I have 

 yet determined; but their arrangement in quartettes in the first 

 two or three whorls seems to be certain. 



The bearing of these observations upon the question of the 

 affinities of Corymorpha is important. The young regenerated 

 hydranth of G. palma is essentially identical in form with young 

 hydranths of Pennaria and Tubularia* Each possesses a flask- 

 shaped body, a whorl of filiform proximal tentacles and one or 

 more quartettes of capitate distal tentacles. This agreement 

 supports the view of Schneider, which has been adopted by 

 Calkins ('99), and with which I am in accord, that the Tiihn- 

 laridce and PennaridcB should be united to form but a single 

 family, since the capitate or filiform character of the distal tenta- 

 cles offers hardly sufficient ground on which to base a distinction 

 between families. The Tubularidce as here used include both the 

 Gorymorphida and Hybocodonidce of Allman ('71) and later 

 authors. For the Hybocodonidce have been distinguished from 

 the Tubularidce only by the possession of free swimming medusae, 

 those of the latter family being permanently attached. The dif- 

 ference is of minor importance. Moreover, Tubularia couthouyi 

 and Corymorpha palma should hardly be placed in distinct fam- 

 ilies on account of differences in extent of perisarc, and character 



*With the uncertain Vorticlava Alder and Aeharadria Wright sis 



the gonosome and are probably immature forms. 



