MORPHOLOGY 327 



which carry male gonophores. The specimen is right-handed and would have sailed to the left of the 

 down-wind direction. It must be the specimen caught during Cook's first voyage, on 7 October 1768, 

 south of the Cape Verde Islands, in latitude 9 42' N., and figured by Parkinson. 



Huxley's (1859) observations, repeated from his 1855 communication, and his review of previous 

 work are most useful. He longed for a Caliph Omar to straighten out the systematics, which, he 

 complained, gave him an unpleasant vertigo. He made no proper distinction between palpons 

 (' hydrocysts ') and gastrozooids, and did not adequately describe the gonodendra ('gonoblastidia'), 

 but to Huxley we are indebted for the first figure of a larva of Physalia. 



Haeckel (1888) in his famous 'Challenger' report had to admit that physalids required a far more 

 accurate anatomical examination than had been employed hitherto. He listed four genera and eleven 

 species. He thought that there were four groups of species of Physalia, two with crested and two 

 with non-crested floats. In each of these divisions there was a group with one and a group with 

 several large main tentacles. I believe that all Physalia are crested in life, and that when fully grown 

 they all have seven or more large tentacles. I have examined hundreds of living specimens of all 

 growth stages in the Canary Islands, as well as preserved specimens from all oceans. I see no grounds 

 for suspecting that there may be more than one species, which should bear the name Physalia 

 physalis (L.). My field studies of Atlantic forms have provided answers to the two questions, which 

 needed answering, before coming to this conclusion; (1) Are some of these animals crested throughout 

 life while others are not, or is this only a matter of age and of degree of temporary inflation or deflation 

 of the pneumatocyst? (2) Do successive growth-stages show (a) differing numbers of large (main) 

 tentacles, and (b) a filling-up of the gap between the oral and main series of cormidia? The facts are 

 that all Physalia are crested, that the number of large tentacles increases with age, and that the gap in 

 question becomes relatively smaller and less conspicuous as growth proceeds. 



Chun (18976) and Bigelow (191 1) both held the view that there were two and only two species, an 

 Atlantic and an Indo-Pacific. But Bigelow assumed that, because up to that time none of the Pacific 

 specimens examined had more than one large tentacle, it was evident that this was the final stage of 

 development and that therefore the fully grown Pacific forms resembled an immature stage of the 

 Atlantic Physalia. This argument has since been shown by Kawamura (19 10) not to hold good. He 

 found in fact later stages of development with more than one large tentacle. 



The figures of Physalia drawn by Haeckel (1888) for pi. xxvi of his ' Challenger ' report give no hint 

 of the great complexity of budding. The only one of these figures which appears reliable is number 4. 

 It is not a lateral view as stated, but a view from above of a right-handed specimen.* It shows very 

 well the general shape of a relaxed specimen, although the basal internode is hardly visible, and the 

 dorsal apical pore and ventral gas-gland cannot be seen. The protozooid is really in a terminal position 

 (PI. XI, figs. 1-3) and not in the subterminal one shown by Haeckel. Fig. 3 of Haeckel's plate showing 

 a young specimen is misleading, because it shows the protozooid as having an ampulla and a gono- 

 dendron, structures never found in that position at any stage, and does not show the oral zone of 

 appendages (see page 337). In fig. 3 (to which I refer again on page 339) and in fig. 6, Haeckel has 

 simplified the structure of the gonodendron. Also the ampulla is always united along most of one 

 side to the tentacle by the muscular fold which runs down the length of the tentacle. Haeckel's fig. 7 

 is valueless and fig. 8 is not at all accurate; fig. 5 is typical of a rather moribund specimen. I mention 

 these points not to detract from the great value of Haeckel's pioneer work, but to help those who are 

 not familiar with Physalia. Haeckel made the most of the specimens of the relatively few good species 



* Brandt's generic name Alophota, used by Haeckel in explanation of the figure, refers to his idea that there were species of 

 Physalia with no crest. 



4-2 



