igi5-] Fauna of the Chilka Lake : Coelenterates. 101 



slightly everted, they tend to grow upwards rather than outwards and so to be 

 folded against the outer margin of the undivided part of the arm. Their ectoderm 

 thus comes in contact with the ectoderm of that part. Ectoderm fuses with 

 ectoderm and is invaded by mesogloea, but as the folding is a simple one no new 

 channels are left open. The characteristic arm of the Triptera is thus produced, 

 formed in its distal region of three lamellae meeting in a vertical line and having a 

 >~ shaped cross-section. 



Another point that may be noticed is the large size of the gastric filaments in 

 the young medusa, in which they are actually as large as — relatively of course much 

 larger than — in the adult. 



Acromitus rabanchatu is a sluggish medusa usually seen on the surface with its 

 main axis nearly horizontal. Its pulsations are slow and feeble. Probably the fixed 

 stage occurs on rocks or weeds near the south end of the lake, where the young were 

 found in April, July, September and November, but not in January or February. 

 Small copepods were noticed in the stomach of the young. The stinging-cells have 

 little or no effect on the human skin. Personally I could detect none. 



The most striking point in what we ascertained as to the biology of this medusa 

 is the effect that an irruption of fresh water has on its habits and physiology. We 

 noticed that medusae were absent from the surface of the northern part of the main 

 area of the Chilka Lake for a considerable part of the year in which they were fairly 

 common in Rambha Bay, although the prevailing wind had a tendency to drive 

 them northwards. The season at which we did not find them on the surface off 

 Barkul and Nalbano was that at which fresh water, which never penetrates fully 

 into Rambha Bay, was prevalent in the northern parts of the lake. At this season 

 our nets often brought up specimens of Acromitus from the bottom; they seemed to 

 be unusually sluggish, to have unusually flat disks and long arms ; but we did not 

 notice anything very definitely peculiar. By a fortunate chance abnormal meteoro- 

 logical conditions made it possible to make a much more definite observation in the 

 Ennur backwater near Madras in January, 1915. At that time, at a season at which 

 the weather is usually dry, heavy rain had fallen and the specific gravity of the 

 water in the upper reaches of the lagoon had sunk, probably quite suddenly, at least 

 as low as i-ooi. No medusae were seen on the surface, but every haul of the bottom- 

 nets brought up specimens ; in one case as many as twenty in a haul. At first sight 

 they appeared to be dead ; no movement of any kind could be detected and the 

 circular muscles of the disk were uncontracted and flaccid. The disks were so flat, 

 owing partly to the condition of the muscles but mainly to an actual shrinkage of 

 the jelly, that the specimens were recorded provisionally as representing either a 

 distinct species or a phase of A . rabanchatu in which the disk retained the post-larval 

 form ; the arms, in consequence of the shrinkage of the bell, appeared to be excep- 

 tionally long. That the medusae were not dead was proved by two facts— they 

 exhibited no signs of decay and the spots on their umbrellas were clear and well- 

 defined. The latter fact is particularly important, because in medusae of this species 



