C 445 ] 



3tio. Now by the table, looo cubic inches at 32°, barometer 

 29,3 weigh 322,1 grains; and confequently 949 cubic inches 

 fhould weigh 305,67 grains. Therefore the error of Sir George 

 Schuckburg's coefficient is exadily 305,67 — 302,89 = 2,78 grains 

 in defedl ; therefore his coefficient retrenched too little, and ought 

 to be higher. It is needlefs to examine General Roy's coefficient 

 at this temperature, though it being higher (2,51) it muft be lefs 

 defedive. 



The coefficient wljich agrees beft with this experiment is 2,84. 



The eafieft method of finding how many meafures, 1000 or 

 any other meafures of air gain by any temperature above 32®, is 

 to multiply 1000, or fuch other meafures, into that fundamental 

 difference, and divide the produd by 400, that being the reci- 

 procal of the coefficient — '1-, which is that I employ ; as ap- 



1000 



plying more exadly in general than any other, though in fome 

 cafes others may be more exadt. 



Second Experiment. 



Barometer 28,77 p 1°. According to Sauffure, hygrometer. 

 Thermom. 68*^ 3 p. 284, a cubic foot of dry air weighs 751 

 grains, barometer 27, thermometer 16?. ' 



Vol. VIII. 3 K Or 



