THE PIMELODUS OP THE VOLCANOS. 213 



lation either from scorching or from the effects of 

 the hot water with which it is discharged. Baron 

 Humboldt states, that in turning over the records 

 kept by the small villages in the vicinity of Coto- 

 paxi, he found mentioned, that, on the lands of 

 the Marquis Selvalegre, so large a quantity was 

 thrown, that a putrid odour was spread over the 

 country. The almost extinct volcano of Imbaburu, 

 in 1691, discharged thousands over the plains sur- 

 rounding the village of Ibarra, and to the miasmita 

 which occurred from them, fevers were attributed ; 

 and from another volcano, in 1698, thousands were 

 also thrown, encased in argillaceous balls. Humboldt 

 is of opinion that these volcanos contain subterra- 

 nean lakes, from whence the supply is afforded, the 

 numbers in the little rivulets around being com- 

 paratively small ; he adds, many of these rivulets 

 may communicate with these subterranean caverns ; 

 and, that the first Pimelodi which have stocked 

 them must have ascended against the stream. In 

 supposing this to account for the numbers destroyed 

 by eruptions, we must conceive their producers to 

 be very abundant. Does the spawn and the young 

 also, in part at least, afford nourishment to the 

 adults ? or in what manner does a fish supposed to 

 have been indigenous to the open streams support 

 itself in these subterranean abodes ? or has it been 

 the reverse, and have the " waters under the earth" 

 supplied the comparatively few specimens which 

 exist in the small streams ? We do not gather from 

 the text that Humboldt saw specimens of the fish 



