3o8 DISCOVERY REPORTS 



After another day there was practically none left. Agassiz and Mayer (1902) recorded traces of this 

 blue colour in their figures of young individuals and also a plum-coloured pigment at the pore end. 

 I have seen traces of this pigment too. 



The first occasion when larvae were collected at Miami was on 21 October 1957. No more were 

 seen until 8 November, and thereafter on 11, 13, 15 and 21 November; on the last date the first large 

 accumulation of Physalia specimens was found. Further collections were made on 9, 12, 16 and 17 

 December and the last and most valuable, already mentioned, on 20 January 1958. The lengths of the 

 larvae are given in Table 1 . 



Table 1 



Date Larval length [mm.) 



8 November 1957 Under 30 



11 November 1957 From 5 to 8, but mostly between 20-40 



13 November 1957 From 19 to 59, but mostly between 30-50 



21 November 1957 From 23 to 150, but mostly between 23-150 



17 December 1957 Probably less than 15, mostly between 60-80 



20 January 1958 From 1-25 to 5 



Miss Dodge informed me that, though the large forms seemed more prevalent, the small ones could 

 be found on certain (not all) days, and in large numbers if at all. 



Physalia feeds on flying fish, mackerel and other surface-swimming fishes which collide with the 

 tentacles, become immobilized by the nematocysts and are drawn up to the hundreds of feeding 

 polyps, which lie on one side of the under-surface of the float. R. P. Bigelow has described (1891) 

 how he was able to keep alive in large ' hatching-tanks ', and study for a week or more, numbers of 

 Physalia taken in the summer of 1889 in Vineyard Sound. His short paper is well worth reading. He 

 reported that the pull of a live fish on a tentacle caused retraction, but contact — except by the mouth 

 of a gastrozooid — with food or its close proximity caused no response. He noted the contrast between 

 the erected crests — relaxed only for a short time — of animals seen in the Gulf Stream and the relaxed 

 crests of specimens kept in the Woods Hole tanks. By directing a stream of air from a bellows on to 

 relaxed specimens he obtained a general contraction of muscles and erection of the crest. He noted the 

 shedding of gonodendra in the tanks and the detachment of gorged animals from the remains of the 

 fish on which they were feeding. He also noticed that the palmar surface of the hand was not sensitive 

 to the sting of nematocysts, which he likened to that of a bee when touched by the back of a finger. 



In 1942 E. W. Gudger gave a resume of recorded observations on the feeding habits of Physalia, 

 from the time of Quoy and Gaimard's 1824 report on the voyage of 'Uranie' and ' Physicienne ' 

 onwards. In 1947 D. P. Wilson published photographs of a 5-5 cm. wrasse, Ctenolabrus rupestris 

 (L.), in the process of being digested by a Physalia about 23 cm. in length. 



Much light is thrown on the mechanisms in coelenterates by a recent paper by W. F. Loomis (1955) 

 on feeding-reactions in Hydra, which are initiated and controlled by a hormone, reduced glutathione, 

 released from the prey itself after penetration by nematocysts. If, as seems possible, the feeding- 

 mechanisms of coelenterates generally is under some such hormone control, it becomes more easy to 

 understand the method in Physalia. Very vast numbers of nematocysts would produce the discharge 

 from the prey of large quantities of hormone and the feeding-mechanism would operate effectively. 

 It may be that the often observed dropping of a half-digested fish is not a sign of repletion but of the 

 autoxidation of the hormone, which, in Hydra, Loomis noted took place rather sluggishly. Since 

 this report went to press, Lenhoff and Schneiderman* (1959) have published the results of experi- 

 ments showing that reduced glutathione does induce a feeding response in Physalia. 



* Tentacles are not lacking, as they state, but are partly separated from the fully grown gastrozooids. 



