1903.] MEDUSJ5 PROM BRITISH COLUMBIA AND ALASKA. 187 



Holl, Mass. Here it was represented by what was thought to be 

 a second species. Two years later Dr. Mayer confirmed this 

 opinion, adding that Prof. Agassiz pronounced the Woods Holl 

 form different from the Pacific form G. vertens, and Mayer (29) 

 proposed the name G. murhachii. 



Agassiz and Mayer (3) described a new species, G. siiavensis, 

 from Suava Harbour, Fiji Islands. Again, Mayer (28) con- 

 cluded that his Guhaia aphrodite was really a species of Gonionemits. 

 The development of Gonionemus seems to indicate that it belongs 

 to the lower Trachomedusae ; there are also other features that 

 put it in the lower Trachomedusse, such as the position of the 

 otocysts, gonads, the marginal welt of nettling-oiguns, and the 

 insertion of the tentacles. Until the life-history of one of 

 the species is better known it is difficult for the present to 

 determine its further position. 



Provisionally, Haeckel's family Petasida?, subfamily Petachnince, 

 with four I'adial canals, otocysts eithei' free on the margin or 

 enclosed, tentacles hollow, is best fitted for its reception. It 

 might then be placed between the genera Aglcmropsis and 

 Gossea. 



With the exception of the number and position of the tentacles, 

 the above characters are so constant in the four species now 

 known that we can look forward to their being found in all true 

 members of the genus. The position and insertion of the tentacles 

 will vary most. 



Only a few Medusfe are recorded, Bathycoclon, Pectantis.kc, in 

 which the ends of some or all of the tentacles are provided with 

 means for clinging to foreign objects. But these are not of the 

 same nature nor in the same position as are the suctorial pads 

 of Gonionemus. Recently, Mayer (27) has found a new Medusa, 

 in which every foui'th tentacle has an adhesive pad that 

 corresponds some^diat with the position of that in Gonionemus, 

 though unlike it in appearance. This Trachomedusa, he thinks, 

 is closely related to Gonionemtts, and has indicated this in caUing 

 it Gonionemoicles. There is no reason why the presence of these 

 pads, if morphologically the same, may not be one of the marks 

 of relationship. The fact that the young of Gonionemoicles 

 geophila have pads on every alternate tentacle, not on every 

 fourth as in the adult, may indicate that in the ancestral form 

 they were so arranged. In this case Medusae having adhesive 

 pads on every alternate tentacle would be more closely i-elated to 

 Gonionemoides than to Gonionemus, or possibly would be inter- 

 mediate forms such as Gonionemus {Cubaia) cophrodite Mayer, in 

 which every other tentacle possesses an adhesive pad. Behaviour, 

 where it is well mai'ked or thei'e are special liabits as in this 

 case, should enter into the characters of the genus or at least the 

 species description. The peculiar habit of swimming to the 

 surface of the water and turning over to float lazily downward 

 is well marked in both G. vertens and G. murbachii, and we may 



