PAEDOPHORE HYPOTHESIS 361 



It is in this context that we may look for the meaning of the tripartite groups of Physalia, about 

 which I said so much earlier on when describing the pattern of budding in the cormidia (pages 328 ff.). 

 Each tripartite group is the homologue of an asexual replicated gonozooid of Velella and of a blastostyle* 

 of Climacocodon, with which I have just compared a siphonophore actinula larva. The difference is that 

 the Physalia polyp, or gastrozooid, is very much elongated, the aboral tentatcles of the Physalia 

 nurse-carrier are absent, and its oral tentacles reduced to a single one, and the single gonozooid 

 complex, the gondendron, has replaced the simple gonozooid or blastostyle. 



If early larvae of Physalia are compared with those of the other two groups of siphonophores, namely 

 the calyconulae of Calycophora and the physonulae of Physonectae, both resemblances and differences 

 will be noticed at once; the large air-sac and tentacle in Physalia, the large nectophore and small 

 tentacle in the calyconula, and the air-sac, tentacle and bracts of the physonula (Text-fig. 30). 



Had it not been for comparison of the cystonect cystonula larva of Physalia with the physonect larva 

 and with the calycophore calyconula, it would have been more difficult to speculate about the ancestry 

 of Physalia. As it is, it seems possible that Physalia' s ancestor arose by neoteny after the first appear- 

 ance of actinula larvae in hydroids, and that it proved advantageous to develop an enormous float, to 

 bud extravagantly and precociously, and to develop bigger and better tentacles with very numerous 

 nematocysts for the capture of the comparatively large fish with which the animal came into contact so 

 effortlessly. 



Thus evolution for Physalia has consisted, I think, partly in its ability through the ages to improve 

 its adaptation to catch and devour fishes at the surface of the sea and to live in a stormy environment. 

 In this respect it may be contrasted with the deep-sea calycophore species of Chuniphyes which have 

 particularly small polyps and no floats. On the other hand the floats of even deep-sea cystonects are 

 large. Physalia is also obviously adapted to drift at an angle of about 45 ° to the right or left of the 

 down-wind direction. Because this angle increases! as me waterline lengthens, it looks as if this 

 adaptation has evolutionary value. If the process of being drifted on to a lee-shore is the analogue of 

 being captured by a predator, the future of Physalia may lie in the direction of improvement in this 

 adaptation. Herein may be part of the significance of the great size of the float in Physalia, and a rapid 

 growth-rate would help in this respect. The apparent neotenic character of Physalia lends cogency to 

 these considerations. 



We do not know that right- or left-handedness is affected by natural selection, and it is difficult to 

 see how one tack can be more advantageous than the other. I do not think that Woodcock's suggestion 

 of advantages to be gained by one tack in feeding is a good one, though the theory of wind-induced 

 vortices no doubt is sound. But it is quite certain that the tack of an individual is fixed from a very 

 early phase of its life-history. The future tack of any individual larva can be recognized already when 

 only the rudiments of secondary buds are becoming visible and the length of the larva is still less than 

 1 -5 mm. The gas-gland itself, even at that age, does not lie symmetrically in the axial line. Because, 

 by custom, we call the meridian in which the tentacle and secondary buds lie ventral, the dorso- 

 ventral plane early comes to lie almost parallel to the surface of the water. The gas-gland itself, while 

 still located at the oral end of the air-sac, then lies asymmetrically, somewhat on the underside. I am 

 of the opinion that right- and left-handedness arises fortuitously when the larva surfaces. Perhaps 

 there is a slight bias of which we know nothing, but the tentacle must come to lie either to the right or 

 to the left, and henceforward the tack is fixed. 



I think the predominance of one tack would be disadvantageous because the maintenance of both 



* Present but not shown in the original figure. 



■j- Woodcock (1956) stated that the young specimens drifted at a smaller angle. I can well believe this, but have not 

 checked it. 



