Tetraplatia — Hand 
345 
ments make their appearance in the ephyra. 
Since Komai suggests the "fins" or lappets 
are the same as those of the ephyra, it seems 
curious that he could so readily dismiss the 
absence of these other structures in this 
animal, Tetraplatia , even though he concludes 
it is neotenic. 
I find that I am in complete disagreement 
with Komai and can only conclude that he 
has not properly interpreted the structure of 
this hydrozoan. One further point might be 
mentioned, however, before the hydrozoan 
nature of this animal is assured and this con- 
cerns the nematocysts. Scyphozoans quite 
generally possess atrichs, although they, as 
well, also possess at least one other type, even 
in the earliest developmental stages (Weill, 
1934). This would seem to be further proof 
that Tetraplatia is a hydrozoan, since its 
cnidom is identical to that of the Narco- 
medusae. There are certain gross characters 
of Tetraplatia which do suggest the Cubo- 
medusae, and some such organism as Tetra- 
platia conceivably could have been involved 
in the early evolution of the scyphozoan line, 
although Tetraplatia is clearly a hydrozoan. 
The problem which now must be solved is, 
"Where among the Hydrozoa should Tetra- 
platia be placed?" I believe Tetraplatia is a 
trachyline medusa and will now discuss this. 
The trachyline medusae are commonly con- 
sidered to be composed of two orders, the 
Trachymedusae and the Narcomedusae. The 
Trachymedusae are defined as possessing free 
or enclosed lithostyles containing endodermal 
lithocytes, a simple bell margin and a mono- 
cnidom of microbasic euryteles. The Narco- 
medusae on the other hand possess only free 
lithostyles, a scalloped bell margin and a 
monocnidom of atrichous isorhizas. The 
Narcomedusae also lack a true manubrium 
and may have peronia and otoporpae. The 
life histories of the trachylines as a whole 
lack a polypoid generation and are said to 
be direct, involving a planula, an actinula, 
and the adult. The gonads of the trachylines 
are sub-umbrellar and occupy a position 
either below the radial canals or in the floor 
of the stomach pouches (Narcomedusae). 
Tetraplatia has a number of characteristics 
which suggest its affinities to the trachyline 
medusae and especially to the Narcomedusae. 
Its swimming plates or lappets can be homo- 
logized with the lappets of Narcomedusae as 
can the adoral stomach pouches which run 
into the buttresses. Its nematocysts are also 
narcomedusan. Its sensory structures, the 
statocysts (enclosed lithostyles), however, are 
trachymedusan. The position of the statocysts 
on the adoral lappet surface is, however, an 
unusual position since all of the trachylines 
have marginal, ex-umbrellar balance organs. 
The position of the gonads, below the stomach 
pouches, is a narcomedusan character unless 
these pouches are actually homologues of the 
radial canals. Its manubrium suggests trachy- 
medusan rather than narcomedusan affinities. 
As an adult animal, Tetraplatia is grossly a 
very differently constructed medusa than any 
known Narcomedusae, in spite of its obvious 
narcomedusan nature. The lack of tentacles, 
radial and ring canals, and the position and 
structure of the sensory clubs make it unique. 
The flying buttresses and the swimming lap- 
pets also are without counterparts. The ex- 
treme reduction of the sub-umbrellar space 
gives this animal its non-medusan, planuloid 
shape. 
Carlgren (1909) was so impressed by the 
peculiarities and unusual structure of tetra- 
platians that he erected a new order of medu- 
sae for them, the Pteromedusae, as a group 
of equivalent rank to the Trachymedusae and 
Narcomedusae. This action has not been 
accepted by other workers as a solution to the 
systematic position of these animals, and, as 
has been remarked previously, Tetraplatia is 
usually considered to be a narcomedusan. 
The Narcomedusae contain several unusual 
forms such as Polypodium and Hydroctena and 
some have very unusual, complicated life 
histories. In general a hydroid phase is absent 
in their life histories, although all known 
seem to involve an actinular larval stage. These 
