ON BALANCES. Gl 



constructed by Mr. Oertling, and made to carry a kilogram 

 in each pan, but differing materially in the length of their 

 beams. 



Long beams have generally been recommended for balances 

 of precision, because the quantity of motion in any point of 

 a lever varies as its distance from the fulcrum ; and, there- 

 fore, the greater the distance of the parts of suspension from 

 the centre of motion of an equal-armed balance, the more 

 distinguishable will be the motion arising from any small 

 difference between the weights compared. On the other 

 hand, there are certain advantages in the quicker angular 

 motion, greater strength, and less weight of a short beam. 

 The larger balance has a beam of ordinary length, being 20 

 inches long and about 2 Ibs. in weight. Its sensibility is so 

 great when in proper condition, that, loaded with a kilogram 

 in each pan, a milligram (equal to y^^- grain, and which is 

 iQOooootn part of a kilogram) added to one of the pans 

 causes a difference in the resting point of the balance of about 

 ten divisions of the index scale. One division, therefore, 

 corresponds with O'l milligram. 



1 7. The other balance has been quite recently constructed 

 from a design by M. Mendeleef, Professor at the University 

 of St. Petersburg. The length of the beam of this balance 

 is only 4f inches, and its weight about ^ Ib. This balance 

 has not yet been sufficiently tested as to its sensibility, but 

 M. Mendeleef claims for his own balance, which served as 

 a model for this balance, that one division of the scale is 

 equivalent to 0*07 milligram only. 



An attempt is here made, and, as it appears, not unsuccess- 

 fully, to overcome the disadvantage hitherto considered to be 

 inherent in a short-armed balance : of its oscillations being 

 necessarily of very small extent, and minute differences in 

 weights compared being consequently not observable. The 

 mode adopted is, as you may see, to employ a microscope 

 with a micrometer for observing the smallest movement of 

 the pointer of the beam over a very finely graduated index. 

 The result has been that, with a kilogram in each pan, an 

 additional weight of a milligram to one of the pans has been 

 found to cause an observed difference of fifteen divisions of 

 the scale. This mode of observing the oscillations of a 

 balance is, however, not new ; it was employed long ago, as 

 I shall presently show. A fuller description of the Mendeleef 



