CORRESPONDENCE. 427 



made nearer or fartlier apart, like the points of sprinfjj compasses. 

 By these means, with the aid of a lens, the ed^es 06 the tool thus 

 constructed are to be made to coincide exactly with the marks desig- 

 nating the length of wire equal to the tenth of a grain. 



Having made this adjustment by the action of the tool, ten pieces 

 of the wire being cut and afterwards weighed against a standard 

 grain weight, if found, too light or too heavy, the screw regulating 

 the distance must be touched so as to cause the distance to be increased 

 or diminished, rendering the cuts larger or smaller. When they arc 

 brought to the weight required, they should, by a delicate assay 

 balance, be tried against each other to ascertain that they are equal 

 in weight to each other. 



Having thus obtained tenths of a grain in weight equal to each 

 other, fifths may be made by the same process and tried against the 

 tenths, tv/o to one, and against each other ; hundredths may be ob- 

 tained by a like process ; for each a tool being requisite like that used 

 for cutting tenths, excepting that it should be smaller in proportion 

 as the lengths required to be cut are shorter. 



The instrument by which these results were obtained has a peculiar 

 capability of reducing the size of the graduatiouR to the limits requi- 

 site to include a greater or less number within any necessary length. 



Suppose it desirable to have as much of a rod as would be equivalent 

 in bulk to a cubic inch of water divided into such degrees as would 

 increase hundredths of a cubic inch. Let a tube sufficiently large to 

 receive the whole length of the rod be at one end recurved at right 

 angles, and terminated in a point with a capillary orifice. Let the 

 other end be furnished with a stuffing-box to receive the rod, making 

 a water-tight juncture. The tube is to be replete with water, the rod 

 entering so as to reach a little beyond the stuffing-box. A mark is 

 then to be made on the rod as close to the box as possible. A light 

 cup being counterpoised on an accurate balance, and a weight equal 

 to a cubic inch of water being placed in the opposite scale, the apex 

 of the rod is to be introduced into the cup while the rod is shoved in, 

 until as much water has been forced into it from the tube as will 

 balance the weight employed as above mentioned. Another mark is 

 now to be cut into the rod close to the box as before. Thus a length 

 of the rod equivalent to the water excluded, and of course equal to a 

 cubic inch of that fluid, is thus indicated to exist between the knife 

 marks. 



The number of ratchet strokes requisite to measure this distance is 

 in the next place to be ascertained and divided by the number of 

 graduations required. The quotient will be the number necessary to 

 make a degree. 



Should the number of the ratchet strokes, when decided as above 

 mentioned, leave a fraction, it is to be gotten rid of by ineans of the 

 contrivance already alluded to for reducing each degree proportionally 

 to any reduction in the whole length necessary to a degree. 



I am willing to put the instrument in question into the possession 

 of the government or of the Smithsonian Institution, on condition that 

 it shall be kept in good order for the purpose of furnishing accurate 

 weights, graduations, or measures of liquids or gases, for the purposes 

 of science and the arts. 



