OX THE DEVELOPMENT OF SYNGNATHUS ACUS. 107 



as the least nutritious of starfish) would not hesitate 

 to allay the wolf at their vitals if a shoal of Syngna- 

 thido? gave them the opportunity. 



The pouch of the male fish is merely external, as 

 in the kangaroo, and is in reality two pouches, 

 formed hy the softer integument of the belly folding 

 around the adhering ova as a still more efficient 

 protection. In the case of our common species, 

 S. lumbriciformis, Kroyer, the ova simply adhere 

 externally, the skin cupping slightly to receive each 

 ovum individually. And here a very interesting fact 

 is worthy of note, in connection with the discussion 

 as to the presence of oil globules in the pelagic 

 ova having relation to their flotation. The presence 

 of oil globules, indeed plentifully, in certain ova, 

 w T ith a little - transparent zona radiata, affixed to 

 objects at low-water, so far settles this question; 

 but the supply of oil globules in plenty to these 

 ova carried about by the fish until incubated, may 

 either be looked upon as settling this question or 

 as raising others ! 



Are these globules of special utility during the 

 formation of the embryo within the ovum, or are 

 the shore species of fishes that deposit their ova 

 in a glutinous mass, adhering to stones or sea-ware, 

 and those deposited on the breast of the parent, 

 both adaptations from an original pelagic con- 

 dition ? If this were the case, the presence of the 

 oil globules would partly explain the fact that some 

 shore fishes, as the rocklings, throw still pelagic 

 ova, while others, like the lumpsucker, throw ova of 

 a more opaque kind, that develop more rapidly, are 

 surrounded by a somewhat tenacious zona radiata, 

 containing sometimes a liberal supply of colouring 

 pigment, and are thus less capable of floating. 

 We find, then, that purely pelagic transparent ova, 

 richly coloured adhesive ova, and ova carried about 

 by the parent, have a supply of oil globules more 

 or less ; so that, whatever their use may be, it does 

 not seem to have any reference to additional float- 



