126 Mr. Drink water on the Preparation of 



that the number "79381 expresses the specific gravity of ab- 

 solute alcohol at 60° F., within a very close degree of ap- 

 proximation. 



My attention was next directed to the best method of de- 

 termining the relative proportions of absolute alcohol and 

 water which exist in legal or revenue proof-spirit. 



There is some ambiguity in the wording of the act of 

 parliament 58 Geo. III. c, 28, which defines proof-spirit to 

 be '•' such as shall at the temperature of fifty-one degrees by 

 Fahrenheit's thermometer weigh exactly twelve-thirteenth 

 parts of an equal measure of distilled water," but the tempe- 

 rature of the water is not stated : there is no doubt however 

 that the temperature of 51° F. was intended to apply to the 

 water as well as to the spirit ; therefore taking water at 51° F. 

 as unity, the specific gravity of proof-spirit at that tempera- 

 ture will be '92308 ; or raising the temperature of both to 

 60° F., the specific gravity will be '91984 — the expansions 

 being calculated from Gilpin^s Tables *. 



It was found by a few preliminary experiments that the 

 specific gravity '91984 would lie between a mixture of 49 

 absolute alcohol + 51 water, and a mixture of 49^ absolute 

 alcohol -I- 50i watei-, all by weight. Mixtures were therefore 

 made in these proportions in an apparatus procured for the 

 purpose, consisting of two light flasks each capable of con- 

 taining about 2200 grains of water. The alcohol and water 

 were weighed separately in these flasks with the greatest 

 care ; after Avhich the flasks were joined without mixing the 

 liquids, the neck of one being ground into the neck of the 

 other for that purpose; the liquids were then thoroughly 

 mixed by transferring them alternately from one flask to the 

 other. The flasks were disconnected when the mixed con- 

 tents became cool, which were then transferred to a clean 

 and dry well- stoppered bottle, and further secured by tying 

 a piece of caoutchouc over the stopper. 



It may also be proper to mention, that when placed in the 

 balance, these flasks were always counterpoised by other 

 empty flasks of the same material, and of very nearly equal 

 size and shape; and to prevent loss from evaporation, a ground 

 glass cap was placed over the mouth of each flask as soon as 

 the exact weight was obtained. The errors of observation 

 could not, 1 believe, in any case exceed one-hundredth of u 

 grain, as the balance used in these experiments was one of 

 Robinson's best instruments, which was previously adjusted 

 by Newman expressly for the purpose ; it was turned with 

 its greatest load by less than one-hundredth of a grain. 

 * Philosophical Transactions for 1794. 



