INFINITESIMAL QUANTITIES RAMSAY. 225 



lamp, which falls on a scale at about 3 meters from the mirror. The 

 filament of the lamp projects its image upon the scale, divided into 

 millimeters. 



The balance is contained in a brass box, in which a vacuum may 

 be made. This box is perforated with two holes upon the lower 

 surface, opposite and below the ends of the beam. There are cemented 

 m these holes two hollow glass-stoppers, into which two tubes of 

 about 3 centimeters' diameter are fitted. In each of these tubes 

 is suspended a very delicate quartz-fiber, soldered to one end of 

 the balance-beam. These threads bend with the greatest ease. 

 There are no pans. Suspended from one of the threads is a small 

 bulb of silica, whose capacity has been ascertained by weighing it 

 filled with mercury. Thus it contams a known volume of air of 

 known temperature and pressure; consequently, the weight of the 

 air is known. A solid counterpoise of silica is suspended from the 

 other thread. 



Now, when the air m the box containing the balance is under ordi- 

 nary pressure, the bulb is balanced by its counterpoise; but if the 

 pressure be diminished, the bulb sinks. If, on the contrary, the 

 pressure be increased, the bulb floats and its apparent weight 

 becomes less. Thus we may, by altering the pressure, add or take 

 away small known weights. It suffices to read the pressure on a 

 manometer, and tlie temperature, so as to have data for the 

 calculation. 



The objects to be weighed are hung on the thread hy small hooks 

 of silica. But before weighing, the sensitiveness of the balance must 

 be determined. This is done by means of a silica rod attached verti- 

 cally to the center of the beam. It is originally longer than is neces- 

 sary in order to be able to draw off small pieces of it by softening the 

 silica with the bloAvpipe. Each time this is done the oscillation period 

 of the balance is determined. Finally traces of silica are volatilized 

 by heatmg in the blowpipe fiame until the oscillation has acquired a 

 sufficiently long period — e. g., an oscillation in 50 seconds. 



The vessel containing the object to be weighed is held in equi- 

 librium by a counterpoise consisting of a hook of silica. With this 

 counterpoise air must be admitted into the box up to a known 

 ])ressurc. We began originally by admitting unpurified air, but 

 were not long in discovering that every trace of dust and moisture 

 must be removed. The air which enters is consequently purified by 

 passing through a column of pentoxide of phosphorus, of soda-lime, 

 and of cotton wool. 



Above a pressure of 150 milhmeters of mercury, currents of air 

 in the box exercise a disturbing influence. It is therefore arranged 

 that the balance shall be in equifibrium at a pressure of about 80 

 milhmeters. To this end we must alter the counterpoise, which is 



