[ 165 3 



pared with the diameter of the large 

 one ; by which difference it may- 

 happen, that the dilation of the air 

 within becomes greater or lefs by 

 warmth, as the fubftance of the 

 glafs be thicker or thinner, and that 

 the capacity of the tube itfelf varies 

 for this reafon. The difference of the 

 remit, however, can be but very 

 fmall from this caufe. 



Thus, by computing the number 

 of fub-divifions to which thefe feven 

 errors may amount, we find them 

 to be 25. But, as all the five mea- 

 sures of airs are let up in the large 

 tube one after another, thefe errors 

 may, if they were all committed, 

 amount to five times this number, 

 or to 125 fub-divifions. 



The miftakes depending from the 

 great tube, or meafure, may alfo 

 amount to feven heads. 



M 3 I. The 



