356 DISCOVERY REPORTS 



leads directly into the stomach. I do not think that this large opening could have been 

 produced artificially in catching or preserving the animal. There is a broad communica- 

 tion between the interior of the stomach and the surrounding medium at the apex in the 

 intact animal. I cannot, however, imagine how the animals may have fed and how they 

 could have remained alive even during a short time with such a wound and a large opening 

 of the stomach on the apex in form of an anus. In specimens of Cassiopea from the 

 coast, which frequently lie on the bottom with the exumbrella turned towards the 

 ground, a similar opening at the apex is often visible, but only in the jelly of the ex- 

 umbrella on the surface, never so deep as to reach the stomach. In other specimens 

 (Sts. 151, 85, 72) a round healed cicatrice is visible lying in a deep crater-Hke groove 

 at the apex of the umbrella (see above p. 355). All this is very strange. 



With regard to the " Stielcanal" Broch in the Michael Sars Report (191 3, p. 7) says 

 that "this apical projection of the stomach into the mesogloea is mostly confined to 

 larger animals and that there is no such canal in the very smallest specimens". In the 

 present series most individuals of the forma dodecabostrycha, some hyacinthina and a 

 few of forma regina have a " Stielcanal", but the specimens in the youngest stage, the 

 'Bigelow' stage, have none. This is very striking indeed, because in this young stage 

 the presence of a " Stielcanal" is to be expected if the animal in fact is derived from a 

 sessile form (Scyphistoma). This is an important and puzzling problem. 



The GASTRIC CIRRI in the large specimens are very long (20-30 mm.) and very 

 numerous (in some medusae certainly many hundreds). 



The GONADS are mostly poorly developed in most hyacinthina and even in many of 

 the large specimens of regina, but well developed in most of the dodecabostrycha indi- 

 viduals. They are easily overlooked from outside, not so much on account of the very 

 dark pigment of the stomach and the ring sinus, but much more on account of the dark 

 chocolate-brown or velvety pigment covering the walls of the subumbrella and of the 

 subgenital porticus. Vanhoeflren writes (1903, p. 23) that in a specimen oi regina 72 mm. 

 in diameter the gonads were not yet developed. The same is the case too with many of 

 the regina stage in the present series. It is strange that the small dodecabostrycha stages 

 generally show the gonads much better developed than the large ones of the regina 

 type. If developed they are greenish or yellowish in colour (see the discussion below 

 on geographical distribution). 



Pigmentation. Maas (1897, p. 64), who regards the colour as a good specific 

 character for regina, speaks about a "leicht rotlicher Ton, der bei entodermalen Teilen 

 in rot oder rotviolett iibergehen kann", but his opinion is based on a sketch from 

 Agassiz (pi. x) which shows the stomach coloured in a very strange and unusual manner, 

 "diflFering interestingly from most other deep-sea medusae" (Broch, 1913, p. 9). It is 

 pinkish on the whole with a dark purple cross and with tracks of pinkish pigment both 

 on lappets and pedalia. In this respect too (see above under pedalia) the figure is 

 certainly not accurate. Vanhoeflten's figure of regina in the Valdivia Report (1903, pi. ii, 

 fig. 6) shows only a small difli'erence in colouring of the somewhat lighter lappet zone in 



