igoy.] Records of the Indian Museum. 143 



The whole perisarc of the hydroid of this species is so delicate 

 that the thecse can only be seen with difficulty even in the living 

 colony ; in preserved specimens their outlines are always distorted. 

 The constant presence among them, in the pond, of the very much 

 stouter and less transparent thecae of the Protozoon FoUicuUna 

 ampulla was at first a source of confusion to me, until I saw both 

 organisms expanded. 



The gonosomes are produced in November, December and Janu- 

 ary. At the beginning of December (1906) the medusae in the pond 

 were still immature, although many of them had almost attained 

 their full size ; towards the end of the same month their gonads 

 were ripe, while at the beginning of the next month only dead or 

 dying medusae could be found By March 17th another brood had 

 reached maturity, having probably been produced by the young 

 gonophores observed on the colonies in January, In March, how- 

 ever, no hydroids were found ; probably they had been killed by the 

 increased temperature of the water, which was perceptibly warm 

 to the hand in the middle of the day. In my aquarium they soon 

 perished unless the glass was shaded from the direct rays of the 

 sun. Neither medusae nor hydroids could be found in the pond at 

 the end of May. 



A peculiarity, which may have been due to the rise in tempera- 

 ture, was noted in the March brood of medusae. Those which reached 

 maturity in December agreed with the original specimens from 

 Ceylon in not having more than two otoliths in each sense-organ, 

 but those taken in March had either three, four, six or eight. When 

 three or four were present, they were arranged in a single series 

 approximately at right angles to the periphery of the bell ; but 

 when the number was six or eight, they formed two parallel series 

 oriented in the same manner. In some instances it was possible to 

 see that the otoliths of smaller series were actually dividing to pro- 

 duce larger ones, the direction of division being always the main 

 axis of the series. The size of the cyst was always larger when 

 six or eight otoliths were present, and in several cases a partition 

 had been formed between the two parallel series, dividing the cyst 

 into two compartments. It was clear, therefore, that the cysts were 

 multiplying by fission. Numerous new tentacles were also being 

 produced, every stage occurring between small rounded swellings of 

 the periphery and fully elongated tentacles. Browne (in Herdman's 

 Ceylon Pearl Oyster Fisheries and Marine Biology, part iv, p. 140) 

 remarked on these swellings and suggested that they were young 

 tentacles, as has proved to be the case. He also observed that 

 while the normal number of otoliths in a cyst was one, two were 

 sometimes present. He thought it probable that this was a case of 

 twinning, but in the light of the observations just recorded it seems 

 more probable that it was one of division. 



I have commented in the preliminary paper of this series on the 

 survival of both hydroid and medusa in small masses of water from 

 which a fresh supply of air was practically excluded. This was as 

 noticeable in the March brood of medusae as in the December one. 



