[ 28 4 ] 



tube exactly divided, and fees at 

 once, without any fhaking, how 

 much of the two airs is deftroyed. 

 If this method is purfued accu- 

 rately, and if the fame interval of 

 time is obferved between joining 

 the two airs and letting them up, 

 the refult will, however, be found 

 different in different experiments, 

 as Dr. Prieflley makes np fcruple 

 to allow. 



I made a great many experiments 

 to find out the reafon of this differ- 

 ence ; but this tafk I leave to Abbe 

 Fontana, who commenced his en- 

 quiries on this fubject prior to me. 

 I will only relate one of my own, 

 which will fhew the reader what 

 refult he may expect from his expe- 

 riments, though performed in the 

 molt regular manner. 



I filled 



