240 NATIONAL ACADEMY OF SCIENCES 



On February 15, 1866, the Secretary of the Treasury Hugh 

 McCulloch addressed a letter to Joseph Henry, then acting 

 President of the Academy, requesting that a committee be 

 appointed to report to the department on the best method of 

 proving and gauging alcoholic liquors, with a view to the estab- 

 lishment of such rules and regulations as would insure a uniform 

 system of inspection of spirits subject to duties.*' Professor 

 Henry accordingly appointed a committee to consider the sub- 

 ject, assuming the chairmanship himself and designating as his 

 associates J. E. Hilgard and M. C. Meigs. At the same time, 

 F. A. P. Barnard, John Torrey and B. F. Craig were requested 

 to prepare tables of standard mixtures of alcohol and water. Dr. 

 Craig was not a member of the Academy. 



The chief difficulty regarding the system recommended by 

 Professor Bache in 1848 was that the Tralles hydrometer, which 

 was the one then proposed, gave percentage in alcohol, instead 

 of percentages in " proof spirit," or a mixture of 50 per cent 

 alcohol and 50 per cent water, upon which all commercial 

 negotiations were based. While the former could readily be con- 

 verted into the latter in most cases, it would lighten the labors 

 of the inspectors if their hydrometers gave readings in proof 

 spirits. Tralles' hydrometer, furthermore, was not adapted for 

 quick observations within one per cent, which it was necessary 

 should be recorded, on account of high duty; or for gauging 

 large quantities of spirits out-of-doors in inclement weather, or 

 under other unfavorable circumstances. It was also found that 

 the tables used by the Treasury Department were not entirely 

 correct. 



In view of these circumstances, the committee set itself the 

 laborious task of finding a more convenient hydrometer, and of 

 preparing new tables. Its report was submitted on July 21, 1866. 

 The recommendations were that following the custom of the 

 trade, the strength of distilled spirits should be estimated accord- 

 ing to their equivalent in proof spirits, and be expressed in 

 terms of percentage of proof spirits rather than by the use of the 



"Rep. Nat. Acad. Sci. for 1866, p. 18. 



