122 



THE BALANCE. 





FIG. 90 (an. proj. \ reed size}. 



suspended from the frame (fig. 35) by a hook made of wire, of'suSl 

 cient length to keep the scale-pans about 1 or 2 cm from the foot of 



the support ; for some purposes it must 

 be higher, and the hook will have to be 

 shorter. Suitable hooks should be kept 

 ready ; if thread be used for the suspen- 

 sion, the balance will be unsteady. Scale- 

 pans supported by a single rod, as in 

 fig. 90, have the advantage of permitting 

 a somewhat freeer use of the hands ; but 

 in weighing, care must be taken to 

 place the load not in the middle of the 

 pan, but as nearly as possible underneath 

 the point of suspension. 



In examining a balance the beam is 

 first suspended without the scale-pans. , 

 If the pointer indicates an inequality, 

 the heavier arm is very cautiously cor- 

 rected by filing. Since the balance may 

 be so faulty as to render it necessary to 

 return it to the maker, it is advisable to 

 postpone this correction until the com- 

 plete examination has proved the balance to be good in all other 

 respects. The inequality may in the mean time be corrected by 

 suspending a short piece of thin wire to the lighter arm. 



The beam is now made to oscillate by being pressed down lightly 

 on one side. The oscillations should be slow and continue for some 

 time. If the beam oscillates too rapidly, its centre of gravity is too 

 low ; if it comes to rest too soon, there is too much friction at the 

 fulcrum': the knife-edge is not sufficiently sharp. 



The scale-pans are now suspended, and each is loaded with a few 

 hundred grammes, until equilibrium is produced. When this is the 

 case, the pans with their loads are changed to the opposite ends of 

 the beam : if the equilibrium remains undisturbed, the balance is 

 1 true.' But if one arm is longer, it will sink when the pans 

 have been changed, because a larger weight at the shorter arm, 

 balanced in the first position a smaller weight at the longer arm, 

 while in the second position the larger weight is suspended from the , 

 longer arm. Now add a small weight to the lighter scale until 

 equilibrium is restored. If this weight exceed Os r 'l or 0& ri 2, return 

 the balance ; for it is not advisable to attempt any correction of this 

 fault. But if the balance is found to have arms of equal length, 

 correct it now by filing for any previously detected inequality iu 



