[ i 7 3 ] 



There are two other great fources 

 of errors that may be committed in 

 the examination of any refpirable 

 air by the nitrous teft, which Abbe 

 Fontana has alfo pointed out and 

 corrected. 



The firfl of thefe two refides in 

 the act of mixing the two airs. If 

 the nitrous air is let up into the large 

 tube, in which was already the air 

 to be examined, and the tube put 

 by for a while without fhaking it, 

 to give time to the two airs to incor- 

 porate with one another ; or if the 

 two airs are firft put in a feparate 

 veffel before they are let up into 

 the large tube ; there will fcarcely 

 ever be two experiments correfpond- 

 ing with one another ; the differ- 

 ence will be fo great, and the remit 

 fo uncertain, that it may amount 

 in one experiment to an error of 



' fifty 



