FLOATATION 313 



After this larval stage has been passed through (Text-fig. 3) the gas-gland becomes a flattened disk 



on the lower part of the enlarged section of the float (PI. VIII, figs. 1,5). I made the following 



rough measurements, under one and the same set of conditions, of fourteen live specimens, whose 



horizontal length overall, omitting appendages, ranged from 35 to 60 mm. (see Table 3). 



Table 3 



Float-length Gas-gland diameter 



(mm.) 



5° 

 5° 

 5° 

 55 

 60 

 60 

 60 



(mm.) 



10 

 10 



15 

 12 

 10 

 11 

 13 



Measurement of other specimens from various localities after fixation in formalin, some with and 

 some without relaxation in magnesium chloride, are given in Table 4. 



Table 4 



Float-length Gas-gland diameter 



(mm.) 



5° 



53 



65 



85 



92 

 140 

 188 



(mm.) 



9 

 11 

 12 

 16 

 18 

 35 

 45 



These data are shown in the form of a graph in Text-fig. 4, from which it can be concluded that 

 Quatrefages's huge specimen may have had a gas-gland of diameter exceeding 70 mm. The largest 

 known to me is that of a ' Discovery ' specimen from St. 3173 (42 35' N., 1 1° 35' W.), whose gas-gland 

 measures 52-5 mm. Haeckel (1888) thought the gas-gland could reach a diameter of 100 or 200 mm. 



Quatrefages (1854) described how a Physalia (right-handed and about 25 cm. in length) that he had 

 been investigating at La Rochelle suddenly lost all its air and collapsed; and, having chanced to 

 glance at it again a quarter of an hour afterwards, he was surprised to see it distended again, although 

 a little smaller than before. He concluded that it must have taken in air through the pore. Haeckel 

 (1888) stated that, by compressing the float voluntarily, the animal could extrude the included air 

 through the apical stigma and sink down. After a short time had elapsed, he said it could rise again, 

 secreting a great mass of air by the pneumadenia and refilling the float. He said he had often observed 

 this process repeated in December 1866 off the Canary Islands. 



I saw no sign of this behaviour at Lanzarote, nor did I succeed in squeezing the air out, owing to 

 the slipperiness of the animal. Louis Agassiz (1862) who had observed thousands of Physalia alive 

 recorded that he had never seen them emptying their air-sac and sinking under the surface of the 

 water, even in stormy weather. 



On the other hand, young specimens appear to act differently. Eschscholtz (1829) found, on 

 irritating a Physalia five lines long (15-16 mm.), that it suddenly expelled all the air from its bladder 

 and sank. Agassiz and Mayer (1902), who studied a number of young animals between 2 and 4 mm. 

 in length taken by 'Albatross' in the autumn and winter of 1899, among the Paumotus and Society 

 Islands, related that ' unlike the adult, these young individuals possess the ability to sink below the 



