^2° •■ Suppofed Oxygen in Snow. 



In order to confirm my deduftion, I immediately poured the water, in which the fifhes hai 

 died, into a receiver, which prefented a large furface to the air, and, a few inftants afterwards, 

 I threw in a fifh of the fame fpecies, which lived very well in it, and might have remained as 

 long as I chofe. It is, then, an inconteflable truth, that the mere want of oxygen was the 

 caufc of the death of the two fifh thrown into the fnow water in the firft experiment; for it 

 is very evident, that when the fnow water was placed in a fituation to rc-abforb the oxygen, of 

 which it had been deprived by its congelation, it became as capable of maintaining the life 

 of fi/hes as any other water. 



Butitfeems certain, from feveral experiments, that fnow water re-abforbs oxygen from the 

 atmofphere more flowly than other waters which were deprived of it. I have before remarked, 

 that fnow water after having been expofed during fixteen hours to the air, contained (o little 

 6xygen, that it could fcarcely fupport the refpiration of afmall fifh for an hour ; whereas I knew 

 from other trials, that waters commonly contain a fufficient quantity of oxygen for the refpira- 

 tion of a fma'l iifli in the fame circumftances for feveral hours. It feems proper, therefore, 

 to conclude, that the water by its converfion into fnowlofes part of its difpofition to abforb 

 the oxygen of the atmolphcre. But in order to determine this more accurately, I chofe to 

 make fome trials. 



I exhaufted all the oxygen contained an two pounds of well water, "by the refpiratlon of a 

 iifli which was kept in the water included in a narrow-necked bottle until it died. When the 

 fifh was dead, I cleared the bottle very exadlly of the whole of the oil, and poured the water 

 into a veffel with a large aperture, in which I kept it expofed to the air for fixteen hours. I then 

 returned the water into the fame bottle, introduced a fifli of the fame fize and kind, and 

 immediately covered it with oil. At the end of four hours the fi(h was ftill alive, but, after 

 half an hour longer, it died in convulfions, in the fame manner as fifhes ordinarily die for want 

 of air. Well water, therefore, though totally deprived of oxygen like fnow water, that is to 

 iky, to the point of being incapable of maintaining the refpiration of fifhes during the fame 

 time, did re-abforb more oxygen than fnow water. 



This cannot be attributed to the quantity of foreign matters ufually contained in fnow water, 

 en the fuppofition that they either hinder the water from re-abforbing oxygen, or the fifli from 

 extraiting it by refpiration ; for I have found that fiihes live for feveral hours in turbid water 

 expofed to the fame circumftances as the fnow water : and I have alfo proved that this water, 

 after having been deprived of oxygen, re-abforbs it in lefs time than fnow water. 



Nevcrthelefs, fnow water, after a long'interval of time, becomes again charged with all the 

 oxygen it can contain, and becomes capable of maintaining the refpiration of fifh, like 

 every other wates. In the month of September, I kept clear fnow water for five days in a 

 bottle ; and in order that the oxygen might infinuate itfelf more eafily, I firft filtered it through 

 a paper, and afterwards agitated it every day in the receiver. This water fo treated, abforbcd 

 as much [oxygen as it was capable of holding ; for when it was introduced into a fmall bottle 

 with a narrow neck, it fupported the life of a filh for nine hours, which was introduced and the 

 the wate covered with oil. 



But 



