66 LECTURES TO SCIENCE TEACHERS. 



Gauss's, but which was first invented by Le Pere Amyot, 

 and is now more generally used in England and Germany, is 

 that of alternation, or first weighing the two bodies against 

 each other, and then repeating the weighings after inter- 

 changing the weights in the pans. By this second method, 

 no counterpoise is required, and half the difference between 

 the two mean resting-points of the index pointer of the 

 balance shows the difference of the weight of the two bodies 

 in divisions of the scale. 



As the value of a division is continually liable to variation 

 according to the condition of the balance, the state of the 

 atmosphere, the weight in the pans, &c., it is necessary, for 

 attaining very accurate results, to determine the value of a 

 division for each comparison. This is done by an additional 

 weighing after a very small balance weight, the value of 

 which is exactly known, has been added to one of the pans, 

 so that its effect on the reading of the index scale may be 

 ascertained. 



28. The most accurate mode generally adopted for noting 

 the results of the oscillations of the balance by observing the 

 movements of the pointer over the index scale, is as follows. 

 We will first take it as illustrating Gauss's method : 



In weighing two bodies A and B against each other, place 

 A in the left-hand pan x, and B in the right-hand pan y, and 

 set the balance in motion. The telescope should be previously 

 adjusted to the index scale on the left-hand side of the 

 balance, so as to enable the observer to see the effect of the 

 weight of A against B. The first turn of the pointer is 

 always to be disregarded, and the readings of the index scale 

 at the next three turns are to be noted. Then stop the 

 balance. The reading at the third turn of the pointer, and 

 the mean of the two readings at the second and fourth 

 turns, are taken as the extreme readings, or the highest and 

 lowest, and their mean is the resting point of the balance. 

 This constitutes one observation. 



The second observation, which completes one comparison, is 

 made by interchanging the two bodies A and B, by moving- 

 each to the other side of the balance, when similar readings 

 are taken of the weight of B against A. As before stated, 

 half the difference between the two resting points of the 

 balance shows the difference of weight of the two bodies A 

 and B in divisions of the scale. An additional weighing is 



