OF NATURAL HISTORY. l2^ 



The refpiratlon of air is not only neceflary to the exift- 

 ence of land-animals, but to that of fishes ot every denomi- 

 nation. Coetaceous fifties, or thofe of the whale -kind, ref- 

 plre, like man and quadrupeds, by means of lungs ; and, of 

 courfe, they are obliged, at certain intervals, to come to the 

 furface, in order to throw out the former air, and to take in 

 a frefh fupply. 



Inflead of lungs, the other fpecies of fiihes are furniihed 

 with gills, through which they refpire both water and air ; 

 for air is univerfally difFufed or mixed with every portion of 

 water. When a free communication with the external air is 

 prevented by ice, or by artifice, liflies immediately difcover 

 fymptoms of uneafinefs, and foon perifli. TElian informs us, 

 that, in winter, when the river Ifter was frozen, the fifhers 

 dug holes in the ice ; that great numbers of fiihes reforted 

 to thefe holes ; and that their eagernefs was fo great, that 

 they allowed themfelves to be feized by the hands of the 

 fifhermen. Rondeletius made many experiments on this 

 fubjecl. If, fays he, fifhes are put into a narrow-mouthed 

 vefTel filled with water, and a communication with the air be 

 preferved, the animals live, and fwim about, not for days and 

 months only, but for feveral years. If the mouth of the vef- 

 fel, however, be fo clofely fhut, either with the hand, or any 

 other covering, that the pafTage of the air is excluded, the 

 fifties fuddenly die. Immediatsly after the mouth of the 

 veffel is clofed, the creatures rufh tumultuoufly, one above 

 another, to the top, contending which of them fhall foonefl 

 receive the benefit of the air*. In the fliallow parts of riv- 

 ers, when frozen, many fifhes are found dead. But, when 

 parts of a river are deep or rapid, the fifhes fly from the ice, 

 and by this means avoid deftruflion. 



Thefe^ and fimilar experiments, have been repeated by Mr. 

 Willoughby, and many other authors 5 and they have uni- 



* Rondeietius, lib, 4. cap. 9. 



Q 



