BEHAVIOUR 377 



another forming a seam. Thus the whole fish, or a large part of it, becomes enclosed in an improvised 

 stomach, or extension of the coelenteron. Fish in various stages of digestion were collected and pre- 

 served on the Canary expedition. The largest, which was an unidentifiable specimen 9 cm. long, had 

 been caught by a fairly small Physalia (gas-gland diameter=i cm.). Another specimen collected by 

 the author is the shapeless lump figured by Hardy (1956, page 120). It is not known how much of 

 a fish is usually digested and how much rejected. A partially decomposed fish might serve as bait to 

 lure other fish into the tentacles, but this is a matter for speculation. 



No original observations were made on the remaining appendages. According to Brooks (cited by 

 Steche, 1907), the nectophores can perform pumping-movements. The author has observed necto- 

 phores in Forskalia pulsating while little more than bell-buds. Like immature heart muscle, the 

 subumbrellar muscle of medusae is active long before its activity is required. Steche's suggestion that 

 the nectophores come into their own only when the whole gonodendron is shed is therefore still 

 tenable. The function of the nectophores, according to Steche, would be to keep the detached gono- 

 dendron in motion and well-oxygenated. It is certainly hard to see what function they could have 

 while the gonodendra are still attached to the parent colony. 



5. Co-ordination 

 For the most part, the appendages behave independently. The capture of prey by one tentacle does not 

 visibly affect the activities of other tentacles or of the gastrozooids. Although the gastrozooids apply 

 themselves eagerly to any object they encounter, the presence of a fish struggling in the tentacles a few 

 centimetres away causes not the slightest alteration in their random movements. Whether this is still 

 true when some of the gastrozooids have begun to digest the fish, is not known. The rolling behaviour 

 of the float is not affected by the activities of the tentacles or gastrozooids. It continues without 

 interruption during feeding. In fact, Physalia usually acts like an assemblage of autonomous indivi- 

 duals. The behaviour of the individuals is linked in terms of functional significance but not in terms of 

 direct nervous integration. 



There is, however, one response in which nervous integration can be seen. When an unusually 

 strong stimulus, such as firm pinching with forceps, is applied to the float, a general contraction 

 follows immediately, or so quickly that no time lag is visible to the naked eye. This contraction involves 

 a sudden shortening of all or a proportion of the appendages. Appendages near to and remote from 

 the site of stimulation are affected equally and, as far as can be seen, simultaneously. The stems of the 

 gonodendra also shorten. The contraction is brief, and the appendages soon fall back into their 

 autonomous activities again. 



This response has the appearance of a nervous through-conduction reaction. Parker (1932), in his 

 valuable physiological study of the nervous system, overlooked this reaction. He also overlooked the 

 earlier account of Bigelow (1891). In it, Bigelow describes how squeezing the float or applying some 

 strong stimulus to the base of the tentacles may evoke a general contraction of the colony. In natural 

 conditions, it would seem likely that very turbulent surface conditions or abrupt collision with any 

 floating object, such as a clump of Sargasso weed or another Physalia would evoke this response. This 

 might be of value in preventing tangling of the tentacles. 



The co-ordination observed is of a most elementary type. It is interesting in that it points to the 

 existence of nervous communication between the different members of the organism. A somewhat 

 similar response takes place in Alcyonium (Horridge, 19566), where, likewise, the response is appa- 

 rently protective in nature. In the physonectid siphonophore, Forskalia, the author has observed 

 similar violent overall contractions of the siphosome appendages following abrupt stimulation of the 

 stem, a closely comparable situation to that found in Physalia. Here, however, a more elaborate form 



