l8 GF FISHES IN GENERAL. 



The gills, it would appear, have a pov'er of abforbing 

 the air, without giving admiflion to the water which is 

 received into the mouth, and ejected by the branchiae, at 

 each TTiOvement of the opercula which cover them ; but 

 in what manner the fubtile feparation is performed, no 

 naturalifl has ever ventured to declare. As the air and 

 water pafs quickly through the gills, without any ap- 

 par.-nt effort to feparate them, it is probable, that only 

 a fmc-.ll quantity of the former is abforbed, but that, by 

 the frequent tranfition cf the water, a fufficient quantity 

 of it rrav be admitted to fupply animals whofe blood is 

 naturally cold, and not in great abundance *. 



But, however f -all a quantity of air naay fuiTice to 

 fupport the life of a filli, fome portion of it is abfolutely 

 neceiTa'y to every living being. Some have been fliut 

 up in a narrow mcutlied veffel, and have lived there for 

 feveral years j but the air is ro fooner excluded, by 

 flopping up the opening, than the animals are fuffocated 

 in the courfe of a few minutes. When this experiment 

 •was repeated with the vcffel filled half with water, and 

 half with air, after covering the mouth, as the air below 

 began to be exhaufled, the fiflies were feen (Iruggling 

 to rife above one another, and inhate a fmall fupply 

 at the furface of the water ; and the fame confequence 

 will follow, if you fallen down the opercula of the gills 

 with a firing, fo that refpiration cannot be performed f . 

 This probably is the caufe why fo many fiflies are de- 

 ftroycd in rivers by fevere frofl : When, by the congela- 

 tion of the whole furface of the water, the external air 

 is excluded, the animals below muft neceffarily perifh|. 



* Willoughby, lib. i. f Rondclet. lib. iv, 



I Willoughby, cap, v. 



