680 PHILOSOPHICAL TRANSACTIONS. [aNNO 17 QO, 



On Hydrometers. — The readiest way of ascertaining specific gravities, and 

 doubtless the most convenient for public business, is by hydrometers; and those 

 of the simplest construction must be best on the whole, especially if more accu- 

 rate means are kept at hand, to be resorted to in case of disputes. An hydro- 

 meter of glass would be the most certain; but whether it be of that substance, 

 or of metal, it should consist of a ball, or rather bulb, so poised as that a cer- 

 tain part should be always downmost in the liquor, and having a stem rising from 

 it on the opposite part, which would consequently keep upright in using the in- 

 strument. On the size of this stem, the sensibility of the hydrometer chiefly 

 depends. In the old areometers the stem was made so large, that the volume of 

 water displaced between its least and greatest immersions was equal to the whole 

 difference of specific gravity between water and alcohol, or perhaps more; whence 

 its scale of divisions must be very small, and could not give the specific gravity 

 with much accuracy. To remedy this defect, weights were introduced, by means 

 of which the stem could be made smaller, each weight affording a new com- 

 mencement of its scale; so that the size of the divisions on a given length of 

 stem was doubled, tripled, quadrupled, &c. according as 1, 2, 3, or more weights 

 were employed, the diameter of the stem being lessened in the subduplicate pro- 

 portion of the increased length of the divisions. Of late this principle seems to 

 have been carried to excess; the number of weights adapted to some hydrometers 

 being so great as to prove very inconvenient in practice. A mean between the 

 2 methods would certainly be best, which might be suited to our tables in the 

 following manner. 



It is proposed to determine the specific gravity to 3 places of decimals, water 

 being taken as unity: the whole compass of numbers therefore, from rectified 

 spirit of water, at 6o° of heat, would be the difference between .825 and 1.000, 

 that is, 175; call it 220 to include the lightest spirit and heaviest water, at all 

 the common temperatures. Of these divisions the stem might give every 20, 

 and then 10 weights would be sufficient for the whole 220. By making the stem 

 carry 20 divisions, an inconvenience much complained of, that of shifting the 

 weights, would in great measure be avoided; because a person conversant in such 

 business would seldom err to that extent in judging of the strength of his spirit 

 previous to trial ; and yet the stem would not need to be so large, or the divisions 

 so small as to preclude the desired accuracy. In conformity to this arrangement 

 it would be proper, that the weights adapted to the hydrometer should be marked 

 with the numbers of the specific gravity, zero on the top of its stem, without a 

 weight, being supposed to mean 800, and 20 at the bottom of the stem to sig- 

 nify 820, which number the first weight would carry; the successive weights 

 would be marked 840, 86o, &c. ; and the division on the stem cut by the fluid 

 under trial would be a number to be always added to the number marked on the 



