GREAT riPE-Fisir. 329 



editions of Pennant''s Britisli Zoology, the upper figure re- 

 presents the female, and the second figure the male of S. 

 acus. The enlarq-cmcnt on the under surface of the second 

 figure, looking like an elongated fin, marks the situation of 

 the distended pouch of a male. Pennant's third figure is 

 the S. ophidion, and the fourth the *S'. lumhriciformis of 

 this work. Neither S. tj/phle nor S. aquoreus axe figured in 

 the British Zoology. 



At what time or in what manner the ova are transferred 

 from the abdomen of the female to the sub-caudal pouch of 

 the male is, I believe, unknown. 



Mr. AValcott also adds, in his MS. that S. acus begins to 

 breed when only four or five inches long. This I have also 

 obtained proof of ; and although examples of this species not 

 uncommonly occur of eighteen inches long, and Bloch attri- 

 butes to it a length of two to three feet, I have a specimen, 

 four inches long only, a young fish apparently of the pre- 

 ceding year, in the opened abdomen of which the ova, in two 

 small lobes, arc full grown. 



M. Risso notices the great attachment of the adult Pipe- 

 fish to their young, and this pouch probably serves as a 

 place of shelter to which the young ones retreat in case of 

 danger. I have been assured by fishermen that if the young 

 were shaken out of the pouch into the water over the side of 

 the boat, they did not swim away, but when the parent fish 

 was held in the water in a favourable position, the young 

 would again enter the pouch. 



The figures of S. acus and ti/phle are correctly represented 

 by Rondeletius, and the characteristic difference in the form 

 and size of the tubular mouth in each is well preserved, 

 liclow the figure, in that work, of the species now under 

 consideration here, several of the young are represented as 



VOL. 11. z 



