98 On a Proposed Neiv Method of Weighing, 



without modifying the speed of the weighing, and the 

 instrument maker can be resorted to for keeping up the 

 absolute sensitiveness of the balance as often as the quality 

 of the work demands, just as under ordinary circumstances 

 when these shifting scales are not employed. The method 

 of using these shifting scales, indeed, is intended for obtaining 

 more accurate readings of the index in decimal subdivisions of 

 the weight employed as the assay pound, and not in any 

 manner to supersede the necessary renovation and adjust- 

 ment of the knife-edges by the instrument maker. It may 

 be stated that the proposition of movable scales is one which 

 has already been put into practice with complete and 

 satisfactory results.* 



Another form of these shifting scales has been conceived, 

 but has not hitherto been put into actual use ; it seems to 

 promise special facilities for adjusting the index readings to 

 even a greater nicety than that belonging to the series of 

 flat scales above described. Instead of a series of inter- 

 changeable flat scales, a small ivory reel, revolving on 

 a horizontal axis, and retained in any position by a 

 small spring, is fixed between the pillars of the balance 

 support, immediately under the point of the index. The 

 curve of the sides of this reel corresponds to the arc 

 described by the point of the index, and around this curved 

 surface, 72° apart, are inscribed the scales corresponding to 

 Nos. 1 to 5 of the above described series of flat scales. The 

 scale agreeing with No. 6 of that series is engraved in con- 

 tact with scale No. 1, or that which divides the full arc swept 

 by the index into 10 degrees on each side of the zero point ; 

 the ten divisions of No, 6 exactly corresponding with five 

 divisions of No. 1. A series of diverging or scroll lines 

 connects these several scales, and in adjusting this revolving 

 index scale it will be merely necessary to turn it round until 

 that part, of which five divisions are found to be exactly 

 equal to '005 of a grain, is immediately under the index 

 point. 



But there are other points concerning the performance of 

 the balance beam which are of great practical significance. 

 If a beam, after careful adjustment, is exposed to change of 

 temperature, its metallic constituent parts become altered in 

 dimensions by expansion or contraction, and, as change of 



* These shifting scales were made for the writer many years since, by 

 Mr. L. Oertling, of London, who contrived them as a substitute for the 

 proposed reel, as it is described in the following paragraph. 



