THRKE-Sl'INEl) STICIvl.HltACK. 173 



arc quite pacific, appear fat, as if full of spawn, never 

 assume the lirilliant colours of the male, by whom, as far as I 

 have observed, they are unmolested. The bite of these little 

 furies is so severe, that I have frequently known it, when 

 inflicted on the tail, produce mortification, and, conse(|nently, 

 death. They also use their lateral spines with such fatal 

 effect, that, incredible as it may appear, I have seen one 

 during- a battle absolutely rip his opponent quite open, so 

 that he sunk to the bottom and died. Another fact in the 

 history of these interesting little creatures also deserves notice; 

 it is curious, and to me unaccountable. Previously to death 

 they re-assume all their brilliant colours, which they may 

 have lost from defeat; but they are not so clear and distinct 

 as when in the height of their power." 



It was one of the results of Baron Cuvier's more attentive 

 examination of the differences among the species hitherto 

 considered identical, that he was led to separate what had 

 before been regarded as one species of Gasterosteus into 

 three, on the foundation of the different desrrees of arminsr 



'^ Oil 



of the plates which cover the sides. On the authority of 

 such a competent observer this division of species at first 

 met with little opposition, and accordingly it found a place 

 in Mr. Yarrell's "History of British Fishes," where each of 

 these supposed species is individually represented. From the 

 first, however, Mr. Jenyns exj)ressed his doubts, which he 

 stated in his work referred to at the beo^inninsr of this 

 article. Bencwed enquiry has gone on to shew further, that 

 if these three varieties are to be definitely distinguished from 

 each other, several others must fall under the same distinction; 

 since there is no exact number of increase or decrease to the 

 plates on the sides, on which characters it was the definitions 

 were built. And these are not the only parts which are liable 

 to variation, although the others have not been taken into 

 account by writers whose attention has been directed chiefly 

 to systematic arrangement. But if for the sake of simplicity, 

 as well as of truth, we can consider the British sj^ecies of 

 this family that are marked by three or four free dorsal 

 spines, as only varieties of one, we may with confidence 

 affirm that no known kinds of fishes are equally disposed to vary 

 their forms or change their apparent character. I shall best 

 VOL. I. 2 c 



