they nust have other poisonous fishes besides the ones which je have cited. It 

 is thou<»ht that these islands must have the nost nanerous and the most violently 

 toxic fishes of finy place in the Pacific. It is si^aificant thnt in the pre- 

 vious literature oven Americans, ^'jho custonnrily do not eat much fish, have 

 noted the occurrence of poisonous fish in '3amoa. Itany of the poisonous species 

 occur in northern Australia, but. apparently only tetraodonts are f oimd in the 

 southern part. 



In the Atlantic many species of poisonous fish have been reported from the 

 "^est Indies and surrcundinT waters, Poey and many others have -ritten of them. 

 They are perhaps nore numerous than those of the Pacific. 



A nxap has been inserted 5n this renort shovjlnrr the iistribution of poison- 

 ous fishes throu'^hout the world. It shows that these fishes are confined almost 

 entirely to the tropic seas, and they are most numerous around isolated islands 

 far from continentai shores. The ^later in these areas is little affected by 

 drainai^e from the L-tnd, it is poor in plankton, extraordinarily clear, and coral 

 reefs are well developed in it. Poisonous fish seem to be especiallv abuninnt 

 around coral atolls. 



because of the parallels between the distribution of coral reefs and that of 

 poisonous fishes, the vjriter is moved to advance the bold hynothesis th^t there 

 is a connection, indirect if not direct, betrjeen the nematocysts of the coral 

 polyps and the poisonous fishes. In -lapan the onlv poisonous fishes arc a few 

 species of tetraodonts, \'?Hich seem to represent the northern limit of distri- 

 bution. It is interestin-^ to note that in Okinavja Prefecture, where coral reefs 

 are found, quite a fevj of the poisonous species cited in this re^iort occur. It 

 is thcu'^ht that the absence of poisonrus fishes from the Indian Ccean, Red Sea, 

 and ?.ast Indies, where quite a fewi of then should be expected to occur, is due 

 to the fact that in many of these areas the coral does not rjrovj ivell because of 

 the effect of runoff water from the Lnnd. The east coast of -sia is a ^ood 

 example, v.ith a great deal of drainage from the land, an ichthy fauna very 

 different from that of the area covered in this investi,<»ation, and no poisonous 

 fishes, (end) 



84 



