94 THE ROYAL SOCIETY OF CANADA 
pre-existing tentacles, until the fundamental plan is more or less ob- 
scured at maturity. 
There are no otocysts. 
Color. After preservation in formalin the manubrium and ten- 
tacular bulbs are yellowish, the ocelli yellowish red. In life the bell is 
almost perfectly transparent and colorless, the tentacles are cream 
white and the ocelli purple-black; the gonads are in some cases faintly 
greenish, in others white or salmon-color. 
Halimedusa is obviously a highly specialized Anthomedusa, but 
its systematic relationships are not altogether clear, for we find com- 
bined in it the characters of Pandeidæ (1909, 1913), Cytæidæ! (1909) 
and perhaps Williide. The general appearance of the medusa, 
particularly its simple gelatinous peduncle, and oral armature of 
sessile knobs, find such a close counterpart in the Cytæid genus 
Oceania (Mayer, 1910) that at first I thought I was dealing with a 
species of that genus. But in Oceania, as in all Cytæidæ and Bou- 
gainvilleidæ, the tentacles are solid (1909, Hartlaub 1911, Mayer 1910) 
and the ocelli on their inner, axial sides (Mayer 1910, Hartlaub 1911), 
whereas in Halimedusa they are hollow with abaxial ocelli, in 
both which respects they agree with the Pandeide. But no true 
Pandeid, (excluding the Bythotiaridae which I believe deserve a 
separate family), has oral knobs or tentacles. And though one 
species of Calycopsis has similar structures (1913), the general 
organization of that genus, and its allies, particularly with respect 
to the manubrium, gonads and tentacles, differs too much from that of 
Halimedusa to allow of the idea that they are closely related. Finally, 
the structure of the manubrium and gonads of our new genus, particu- 
larly their four radial extensions over the peduncle, and the possibility 
that the sexual products include the bases of the radial canals, suggests 
the Williide, from which family, however, it is separated by the 
simple canals, as well as by the presence of ocelli, and labial knobs. 
We have here one of those cases, so commonly confronting the 
student of Medusz, where a new genus contravenes all family lines, 
as previously laid down, though it is obviously an off shoot from one or 
other family previously recognized, not the representative of a new 
group. 
The hollow tentacles, and particularly their hollow basal bulbs, 
show that Halimedusa is most closely related to the Pandeide in 
spite of its superficial resemblance to some Cytæidæ, e.g., Oceania 
and Turritopsis. And the only important feature in which it differs 
* Hartlaub, 1911, includes Cytæidæ and Bougainvilleidæ in one family: and no 
doubt their members are more closely allied to one another than to the Pandeide 
or Williide. But the distinctions between them seem to me to warrant the retention 
of both families (1909, 1913). 

