applicable to the Gold Bullion Assay. 95 



to heighten or lower the ball with certainty a known 

 number of threads of the screw, or, if necessary, any sub- 

 division, as one half or a quarter of a turn. 



These preliminary adjustments effected, the balance is to 

 be loaded with the full load employed in the actual work, 

 for example, with ten grains in each pan ; and by means of 

 the rider-weight, the two sides are to be brought into exact 

 equilibrium. Ascertain critically the justness of the 

 equipoise, then move the rider so as to make one side of the 

 beam "005 of a grain heavier than the other side, and now 

 observe, by aid of the seconds hand of a watch, the time 

 required for the oscillations of the loaded beam ; observe, 

 also, how many divisions of the index are equivalent to the 

 extra load of "005 of a grain. I will suppose that in this 

 trial repeated experiments show that with the loaded pans 

 for this weight of '005 of a grain the needle sweeps from 

 zero across two divisions of the scale, returning again to 

 zero, and that the vibration is more than sufficiently rapid 

 for our work. We now make a second trial : the ball is 

 raised by two revolutions on the screw, the equilibrium of 

 the loaded beam is adjusted by altering the position of the 

 rider until the needle sweeps over an equal space on each 

 side of the zero mark of the scale, and now, by moving the 

 rider so as to be equal to an added weight of "005 of a grain, 

 we again try the time of oscillation and sensitiveness of the 

 balance : the oscillations have become less rapid in conse- 

 quence of the raising the centre of gravity of the beam, and 

 the beam has at the same time become more sensitive, the 

 '005 of a grain being now represented by a larger sweep of 

 the needle. 



But after two or more trials of this kind we not only 

 arrive at a nicer and still nicer adjustment of the sensitive- 

 ness and time of oscillation, but the comparison of our 

 results affords a valuation of the thread of the screw upon 

 which the ball is raised and depressed, as a means of 

 adjusting with certainty the performance of the beam ; each 

 revolution of the ball, raising it by the width of one thread, 

 is found to retard the oscillation by a certain fraction of a 

 second, and to increase its sensitiveness by a certain fraction 

 of a division of the index-scale per '00 1 of a grain added. 



Patient and careful trials of the kind will enable us to 

 adjust the balance so as to ensure the greatest sensitiveness 

 compatible with its weight and workmanship, or we may 

 compound between sensitiveness and rapidity, sacrificing the 



