114 INTRODUCTION. 



Fish are generally accused of having no attach- 

 ments for their offspring ; they neither construct a 

 nest, nor do they feed or defend their young. There 

 are, however, exceptions, and among these belongs 

 the fish which is known at the coast of Guiana 

 under the name of hassar or hardback {CalUchthys 

 suhulatus*), which constructs a regular nest of blades 

 of grass and leaves, in holes just above the surface 

 of the water, where it deposits its roe and watches 

 it with maternal care until the fry is brought forth. 

 Somewhere else I have related the maternal care 

 which the lau-lau and other species of siluridw 

 show for their young offspring, which swim in 

 shoals about them, and in case of danger enter the 

 large throat of the mother, t These are only single 

 instances ; the far greater number, when once 

 brought into external life, are left to themselves, 

 and perish, devoured by larger fish, reptiles, or 



* I believe that Hancock's CaUicldhys littoralis, or Valen- 

 cienne's C. subulatus, and Margrave's Tamoata, are nearer re- 

 lated than suspected. I am not aware that the Doras in the 

 interior built any nests, and refer therefore only to the Cal- 

 lichthys at the coast regions or brackish waters. 



+ I extract from some manuscript notes of Dr. Hancock's, 

 formerly a resident in Demerara, and which only lately have 

 been lent to me, the following confirmation of the above : — A 

 large gillhakra (a species of salt-water silurus), just taken out 

 of the water, was seen by Mr. Gibbs, and many others at the 

 King's Stelling, to vomit up vast numbers of the young fry of 

 the same species, he thinks three or four hundred, each the 

 length of a finger. The fish had been chased up and taken in 

 shoal water. Something similar has been observed of a cuirass^ 

 another species of sUb-vis. 



