Z^2 On the Hydrometer. 



Thefe examples are, I hope, fafficlent to explain this table. 

 The nfes to which it is applicable, are, that as it affords Jirjl 

 an unequivocal jlandard for every gradation of Jtrength, an 

 hvdronieter can be graduated on a certain principle ; and, 

 lecondly, the errors of thofe that are graduated otherwife 

 may be difcovered bv it, and we fliall no longer be obliged 

 to rely on a mere aflertion, that the inftrumenl is correct, 

 nothing more being required than to compare the fpecific 

 gravity of the fpirit with the indication of Ilrength which it 

 has demonftrated. 



On confidering how far it might be prafticable to firnplify 

 the hydrometer, I perceived that the makers of thefe inftru- 

 ments had fallen into two errors, which had very confidera- 

 bly increafed the neccffity for thofe complex additions with 

 which they are incumbered. 



Firil, conceiving that the infl:rument was to be valued 

 in proportion as it was capable of making minute difcrimi- 

 nations, (or of its fcnfibility as it is called,) they have, by 

 attempts to increafe this beyond thofe limits required either 

 in revenue or in commerce, deftroyed its funplicity. 



In the conltruftion of tliis inftrument, the proportion of 

 the Item to the ball is an important confideration : if the 

 former be too large, the inflrument will not be fufficiently 

 fenfible; and fliould it, on the contrary, be too fmall, the 

 number of gradations of ftrength it is capable of indicating 

 will proportionably be diminidied. 



Thofe who have conftrufted moft of tlie hydrometers now 

 in ufe, have run into the latter error ; the ftem being dimi- 

 nillied to a fize which rendered it necefl'ary to have recourfe 

 to complex additions to fupply the place of this injudicious 

 and unneceffary wafte of the fimple and proper power of the 

 infirument. 



From a great variety of experiments to afcertain the proper 

 proportion between the ball and ftem, I was finally fatisfied 

 that the latter would admit of fuch enlargement as to be ren 

 dered capable of mcafuring upwards of 66 gradations or per 

 centages of ftrength, without any weight; and this fo clearly 

 and diftincliy, as to convince me that its fcnfibility and ac- 

 curacy were ftill fully fuiTicient. 



The fecond error jJludcd to has arifen from an attempt 

 to extend the power of the inftrumcnt beyond thofe limits, 

 which, with propriety, and confiftently with its fimplicity, 

 could be effected. 



An hydrometer being a counterpoife to afcertain the weight 



of a fpirituous liquor, may, with propriety, be compared to a 



6 " pair 



