28 



means of the stereometer, invented by M. Say, and used by 

 Professor Miller. It is described as follows : — " It consists of 

 two glass tubes of equal diameter communicating with each 

 other at the bottom. One tube is graduated by lines traced on 

 the glass ; it is connected on the top to a cup capable of being 

 closed airtight. The body to be tested is placed in this cup. 

 The volume of air in this cup is measured by means of mercury 

 poured into the other tube, by which the graduated tube be- 

 comes a barometric tube. The volume of the weight is shown 

 by the difference of the volume of air before and after the 

 standard weight is inserted in the cup. By means of this 

 instrument the specific gravity of soluble bodies, gunpowder, 

 &c, may be ascertained without injuring them." 



In order, therefore, to ascertain the true relative weights, 

 as well as the actual weights of standards differing in density 

 when weighed in air, it is necessary to allow for the weight of 

 the volume of air displaced by each standard. Thus, in weighing 

 two standards of different material for the purpose of deter- 

 mining their relative weights, it becomes necessary to reduce 

 these weighings to a vacuum by deducting the weight of the 

 volume of air displaced by each of them. 



The correction is to be made according to the formula, thus 

 given by Professor Miller : — " If the weights of P and Q appear 

 "to be equal when weighed in air, then the weight of P — 

 weight of air displaced by P = the weight of Q — weight of 

 air displaced by Q." As a body when weighed in water weighs 

 less than when weighed in air, by the difference of the weight 

 of air and the weight of water displaced by its volume, in the 

 same manner a body weighed in air weighs less than when 

 weighed in a vacuum, by the weight of air which its volume 

 displaces. 



But the weight of air is affected by temperature, aqueous 

 saturation, &c, hence the correction is to be ascertained from 

 the following data : — 



1st. Mean temperature of the air during comparison. 



2nd. The mean barometric pressure reduced to 30° Fahr., and 

 corrected by deducting the pressure of vapour in the 

 air. 



3rd. The density of the metal of which each standard is 

 composed. 



4th. Its co-efficient of expansion. 



5th. The apparent weight of each standard. 

 And the method of computing the weight in grains of air 

 displaced by each standard is by adding the logs, of the follow- 

 ing terms : — 



1st. The given barometric pressure in air, corrected and 

 reduced to 30° Pahr. 



