APPENDIX. 727 



sides of the niduth. The lupostome then elongates. This stage, with two well-developed 

 tentacles and the hvpostome, persists for several days. 



A third tentacle similar to the first two then develops and the three tentacles set them- 

 selves 120° apart, giving the polyp a triradial symmetry. Finally, a fourth tentacle develops 

 and the larva has 4 knobbed tentacles 90° apart. No later stages were observed at RoscofF. 

 France, where these studies were undertaken by Wietrzykowski. 



Genus Thaumatoscyphus Kishinouye, 19 10. 



ThiiuitiatoscyphuSf Kishinouye, 1910, Journ. College of Sci., Tokyo, vol. 27, an. 9, p. 2. 



The type species is Thaumatoscyphus distinctus Kishinouye, from the most northeastern 

 island of Chishinia, Kurile Islands, Japan. 



GENERIC CHARACTERS. 



Stauromedusae closely allied to Halulystus and Stenoscyphus, but with a unitary coronal 

 muscle in the exumbrella; with rudimentary adradial lobes and small, non-adhesive perradial 

 and interradial tentacles.. 4 interradial pits in the subumbrella. Peduncle 4-chambered. 

 8 adradial gonads. Gastric cavit\' as in Eleutherocarpidas. 



This genus is distinguished from all other Stauromedusje by its exumbrella coronal muscle. 

 This structure is so remarkable, being unknown in any other Scyphomedusae, that its existence 

 recjuires confirmation, for contraction in preservation may have produced the furrows which 

 Kishinouye observes and believes to be the outlines of strands of muscle fibers. He cut no 

 sections. 



Thaumatoscyphus distinctus Kishinouye. 

 Thaumatoscyphus distinctus, Kishinouye, 1910, Journ. College of Sci., Tokyo, vol. 27, art. 9, p. 2, plate 1, figs. I ami 2. 



Body goblet-shaped, 30 mm. high. Cal\x 15 mm. wide and half as high as height of 

 entire animal. 8 short, adradial lobes, each with about 40 short, captate tentacles growing in 

 a lanceolate tract on the aboral side of each lobe. The tentacles in the proximal part of the 

 tentacular tract have very large, swollen stalks and degenerate dis:al knobs. These swollen 

 stalks serve as adhesive organs. 



The 8 perradial and interradial tentacles are small, cylindrical, and without well-developed 

 distal knobs, although their ends are captate with a median depression at the tip. These 

 tentacles bear black pigment at their bases and along the median line. They are not adhesive 

 organs. The peduncle is more or less quadrate, about as long as the cal\x and 4 times as 

 long as wide. It is 4-chambered. 



Four deep, interradial infundibula in the subumbrella. The subumbrella is beset with 

 large, spherical, wart-like clusters of nematoc)sts, those near the margin and middle parts 

 of the mesogonia being the largest, and about i mm. in diameter. 



The coronal muscle is a broad, undivided band, the greater part of which is said to lie 

 in the exumbrella beyond the clusters of tentacles. 4 broad but weakly developed per- 

 radial areas of radial muscles extend from the pyloric region through the stomach wall. The 

 interradial muscles are better developed and extend from the aboral end of the peduncle to 

 the bell-marsrin. Each interradial muscle band is divided at its distal end into two short 

 limbs which extend to the bases of the adradial clusters of tentacles. 



The oesophagus is short, somewhat ([uadrangular, and with deep longitudinal folds. 

 The 4 lips are folded. The central stomach-cavity is long and prismatic and there are 8 

 adradial rows of simple, long, gastric cirri. There are 8 adradial lanceolate gonads, each con- 

 sisting of 7 or 8 oblong follicles. The abaxial surface of each gonad is black and can be seen 

 through the translucent wall of the bodv. 



Two specimens found in August, 1903, from Shimushiri, Kurile Islands, Japan. 



Unfortunately Kishinouye appears to have cut no sections and he bases his statement 

 of the existence of an exumbrella coronal muscle upon the presence of annular folds in the 

 external surface of the body-wall. This appearance ma\' well be due to unnatural contraction 

 in the killing fluid. He studied two preserved specimens. Even if this coional muscle does not 

 exist, the medusa may still be called Thaumatoscyphus, for I't is distinguished from Stenoscyphus 

 by its adradial lobes, and from Haliclystus by having 4 subgenital pits in its subumbrella. 



