374 



DISCOVERY REPORTS 



Many observers have commented on the strange rolling-activity and somersaulting performed by 

 the float of Physalia in calm weather. Wilson (1947) suggests that this activity serves to keep 

 the float moist. This would be particularly important in calm conditions in tropical waters. Perhaps 

 a better interpretation of the activity is that proposed by Totton (personal communication). The 

 upright position of the float, with sail ' trimmed to the wind ' is inherently unstable, and can only 

 be maintained when the wind is exerting a certain pressure against the side of the float. If there is little 

 wind, the instability of the float in the sailing posture becomes manifest, and the float collapses on to 

 its side. The float responds to collapse by movements which have the effect of restoring the upright 

 position. In the upright position again, a new attempt at the sailing posture takes place, which, if the 

 wind is still insufficient, results in another collapse. 



I 



Text-fig. 2. A, View of the float from above, to show the angle from which B is seen. The apical pore end (dotted) has 

 been removed at the level indicated by the broken line. The openings around the bulge region and the side lead into the 

 groups of appendages. B, View into the interior of a float from which the apical pore end and the whole saccus has been 

 removed. For convenience, the float and crest are drawn as if they were still supported from within by the saccus, but 

 if an animal were actually opened in this way, the float would collapse completely. Note the vaulting in the roof, forming 

 pockets into which the saccus outpushings would normally fit. 



The author's observations, considered in retrospect, support Totton's interpretation. It was 

 several times observed that a sudden drop in the wind would cause sailing specimens of Physalia to 

 fall over on to the windward side. Specimens were much less apt to fall on to their leeward sides ; to 

 do so would mean hauling a considerable weight of tentacles up out of the water. Bigelow (1891) was 

 the first to draw attention to this fact. The float is precariously balanced, and slight local contractions 

 of the muscular wall can bring about shifting in the centre of gravity and consequent rolling into a 

 new orientation. Rolling occurs mainly about the a-b axis (Text-fig. 1), sometimes about the b-c axis. 

 In the latter case, the apical pore end may rear up high in the air, and a complete somersault take 

 place. These variations are understandable when one considers that all three sides of the float are 

 contractile and that contraction can be localized in particular areas. A considerable range of body 

 form can be assumed, in all of which the pattern of rolling will be different. It was found possible, 

 by electrical stimulation of appropriate regions of the float wall, to evoke local contractions of sufficient 

 power to bring about shifting of the centre of gravity and various kinds of rolling. 



