74 HISTORICAL INTRODUCTION TO CHEMISTRY CHAP. 



Priestley tests the goodness of common air by mixing 

 it with nitrous air. Priestley's method of testing air 

 was described in detail a few years later as follows : 



" I first provide a phial, containing about half an ounce 

 of water, which I call the air-measure. This I fill with air 

 by having first filled it with water, and placed it over the 

 opening of the funnel in my shelf, and when it is filled I 

 slide it along the shelf, always observing that there be a 

 little more air than I want. The phial being thus exactly 

 filled with the air which I am about to examine, and care 

 being taken that it be not warmed by holding in the 

 hand, &c. I empty it into a jar about an inch and a half 

 in diameter, and then introduce to .it the same measure of 

 nitrous air, and let them continue together about two 

 minutes. I choose to have an overplus of nitrous air, that 

 I may be sure to have phlogiston enough to saturate all the 

 common air. If I find the diminution with these measures 

 to be very considerable, I introduce another measure of 

 nitrous air ; but the purest dephlogisticated air will not, I 

 believe, require more than two equal measures of nitrous 

 air." 



"Sometimes I leave the common and nitrous air in the 

 jar all night, or a whole day ; but always take care that, 

 whatever kinds of air I be comparing together, they remain 

 the same space of time before I proceed to note the degree 

 of diminution." 



" When the preceding part of the process is over, I 

 transfer the air into a glass tube, about three feet long, and 

 one-third cf an inch wide, carefully graduated according to 

 the air measure, and divided into tenths and hundredth 

 parts ; so that one of the latter will be about a sixth or an 

 eighth of an inch. Then immersing the tube in a trough of 

 water, so that the water in the inside of the tube shall be on 

 a level with the water on the outside, I observe the space 

 occupied by them both, and express the result in measures 

 and decimal parts of a measure, according to the graduation 

 of the tube " (Experiments and Observations, 1779, IV. 

 Introduction, pp. xxx. xxxii.). 



Using this method, Priestley in 1772 thought he perceived 



