376 DISCOVERY REPORTS 



The most characteristic activity of the tentacles is a rhythmic shortening and lengthening. On the 

 other hand, tentacles, whatever their rate of contraction, can remain completely motionless for long 

 periods. Bigelow was the first to describe the shortening and lengthening of the tentacles. He did not 

 regard it as sufficiently regular to be called 'rhythmic'. Parker (1932), however, found evidence for 

 rhythmicity in the movements, the interval between successive performances being 30-75 sec. In the 

 present investigation a definite rhythm was observed. It was sometimes a very loose one as in the 

 cases studied by Parker, but in many examples, particularly where fresh healthy specimens were under 

 observation, a much more regular rhythm was in evidence. It was found that the rhythm was slower 

 in small tentacles than in large ones, but that in all tentacles, whatever their size, the frequency 

 depended on the general degree of contraction of the tentacle within which the smaller rhythmic 

 changes in length were taking place. In one case a small moderately extended tentacle was timed 

 over 23 min. The contractions followed one another every 10-17 sec - The specimen was then trans- 

 ferred to another tank and, as invariably happens following such disturbance, the tentacles shortened 

 considerably. When the rhythmic activity emerged again in the original tentacle studied, the con- 

 tractions were coming at 5-10-sec. intervals. Gradually the overall length of the tentacles increased 

 again, and the frequency of the rhythm declined. 



Rhythmic shortening and lengthening seems as a rule to accompany gradual changes in overall 

 length and it may, in fact, be the basis for such changes. In this there is a superficial resemblance to 

 the rhythm in Metridium (Batham and Pantin, 1950). However, in Metridium the rhythm is slow, 

 sigmoid and distinct from the facilitated responses evoked by mechanical stimulation, while in 

 Physalia the contraction phase of the cycle is rapid, ' stepped ', and indistinguishable from a mechanic- 

 ally evoked contraction. The ' steps ' are much more noticeable in the contraction of the tentacles than 

 in contraction of the float because the amplitude of the contractions is much greater. Each ' step ' may 

 shorten the tentacle by several inches. Not more than three or four ' steps ' are distinguishable in the 

 contraction phase of the rhythmic movements. The elongation of the tentacle following one of these 

 stepped contractions is smooth and gradual. It is probably passive, being caused by gravity or drag. 

 Endodermal circular muscle is present in the tentacles, but there is no evidence that it assists elonga- 

 tion. It is more likely that it serves for movement of fluids in the coelenteron. 



Rhythmic shortening and lengthening is carried on independently by individual tentacles, its rate 

 depending on their sizes, and states of contraction. We are evidently dealing with a well-developed 

 local action-system, such as frequently characterizes structures with a high degree of autonomy 

 (Pantin, 1952). It was observed in the laboratory that rhythmic movements had the effect of unravel- 

 ling any tangles which had occurred among the tentacles. This, however, may be no more than an 

 incidental occurrence. 



4. The Behaviour of the remaining appendages 

 The gastrozooids show a considerable amount of activity in some specimens. This consists of random 

 ' searching ' movements. On touching a solid object they apply themselves to it, the mouths spreading 

 out over any area up to 1 cm. in diameter. Although Bigelow (1891) stated that attachment to inorganic 

 objects was of brief duration, it was observed in the Canary Islands that gastrozooids sometimes re- 

 mained attached to the polythene wall of the tank for periods of more than an hour. It is, however, 

 too early to say whether or not the feeding response involves chemical as well as tactile stimuli, as 

 in Hydra (Semal, 19546).* 



The appearance of a fish enshrouded by feeding gastrozooids is well-known from Wilson's account 

 {pp. cit.). The edges of the zooid-mouths spread out until they touch and then press up against one 



* It does (see footnote on p. 308). 



