24 ABSORPTION OF 



pletely phloglfticated ; fo that this long fpace of water 

 had been Uttle or no obftrudion to this procefs. 



On the 2 1 ft of Jan. I fet by two quantities of com- 

 mon air, each one meafure, in two fimilar glafs jars, one 

 plunged feveral inches under the water, and the other 

 placed on the flielf in the fame trough, thinking that a 

 difference in the prejfure to which they were fubjeded 

 might make fome difterence in the abforption ; and till 

 the 26th of March that which was on the fhelf was 

 more diminiflied than that which was under the'water, 

 and therefore more comprefled, but on that day they 

 were exactly equal, viz. 0.55. After this that which was 

 funk in the water was more diminifhed than the other. 

 On the 30th of April, that which was funk was 0.48, 

 and that on the fhelf 0.59 ; but on the ift of July, when 

 I put an end to the experiment, the changes were re- 

 verfed again ; for that which was funk was o, 1 7, and 

 that on the fhelf 0.08. 



1 found, however, that dilatation by an air-pump pre- 

 vented the abforption, and comprcfTion by a condenfmg 

 machine rather promoted it. To determine this I Inb- 

 jeded one meafure of common air to the preffure of 

 about two atmofpheres a month, in which time 1 kept 

 another equal meafure dilated about fix times, and ano- 

 ther in a fimilar veffel without dilatation or compreffion. 

 This was then found to occupy the fpace of 0.85, the 

 comprefled air 0.76, and that which had been dilated 

 had undergone no change at all. 



I repeated this experiment on nitrous air from the 

 15th to the a8th of March, when the compreffed air oc- 

 cupied the fpace of 0.47, the dilated was 0.91, and that 

 which had been neither comprefled nor dilated was 0.54. 

 They had all loft their power of affeding common air 

 nearly in the proportion of their diminution. 



The 



