Ch. XXI.] POETUGUESE MAN-OF-WAE. 365 



was red, inflamed, and swollen, and remained so for some 

 hours after the occurrence. 



One circumstance in relation to these large Physaliae 

 . struck me as being very remarkable. Each one as it floated 

 by had beneath it what at first I took to be its mass of ten- 

 tacles and polypites ; but, on more close observation, I 

 found that the appearance was due to a shoal of small 

 fishes accompanying the hydrozoon under protection of its 

 appendages. The fishes were of various sizes, from 2 to 6 

 inches long, transversely banded, and looking in the water 

 precisely like the pilot-fish (Naucrates ductor). There were 

 perhaps a dozen of these accompanying fishes clustered to- 

 gether beneath the bladder of each Physalia. Every Physalia 

 had its cluster ; but this peculiarity was observable, viz. that 

 under small Physalise the fishes were small, while under 

 large specimens they were correspondingly large, being in 

 fact, always proportioned to the size of the man-of-war 

 which they accompanied. Unfortunately I did not discover 

 this curious fact till late in the day ; and when the boat was 

 down in the morning I was unaware of it, or I should have 

 made a point of attempting to secure a specimen of so inte- 

 resting a fish. 



What the relation is which exists between the fish and the 

 hydrozoon I cannot say; but this correspondence between 

 the sizes of the two animals seems to indicate that the fishes 

 do not capriciously select their protecting Physalia. It is 

 known that certain fishes harbour in the threads of the 

 larger Lucemaridse, or umbrella-form Acalephs ; but I be- 

 lieve they have not before been noticed to accompany the 

 Portuguese man-of-war. 



The presence of these fishes also accounted for a remark- 

 able circumstance I had observed earlier in the day. A large 



