applicable to the Gold Bullion Assay. 97 



have thus restored the sensitiveness of the beam we have 

 thereby added to the time required for each weighing : the 

 beam will oscillate more slowly. A better arrangement is 

 that in which a series of scales, any one of which may be 

 easily and quickly slipped into position under the index, is 

 employed. The ordinary scale divides the full sweep of the 

 index into twenty equal divisions, ten on each side of the 

 zero point. If, now, we provide a series of five other scales 

 in which smaller arcs are divided into twenty divisions, ten 

 on each side of the zero point, we can from time to time 

 substitute one or other of these scales for that originally in 

 use, which latter, for distinction, I will call the ordinary 

 scale. When the ordinary scale is replaced by one of these 

 scales, we can weigh with the same speed, and virtually the 

 same sensitiveness of the beam, and with no other dis- 

 advantage than such as may belong to a scale of somewhat 

 closer divisions. 



In these proposed supplementary scales ten divisions on 

 each side of the zero mark are respectively equal to the 

 following proportions of the ordinary scale. 



In scale No. 2, 10 divisions = an arc of 9 divisions on No. 1, the ordinary scale 

 3, 10 = 8 



„ 4,10 = „ .7 



5, 10 ,, = „ 6 ,, ,, „ 



6, 10 „ = „ 5 „ „ „ 



and with this series we can preserve the decimal value of the 

 index readings until the sensitiveness of the beam has 

 diminished by one half. The following example will 

 illustrate the way in which these scales are brought into use : — 

 For some time after the new beam is employed, no measure- 

 able diminution of sensitiveness can be observed ; but 

 eventually the knife-edges become a little impaired, and 

 when critically tried it is found that for -005 of a grain the 

 index sweeps only 4 \ divisions of scale No. 1 ; this is 

 equivalent to sweeping 9 divisions for "01 of a grain, and we 

 can now readily compensate this diminished sensitiveness by 

 replacing scale No. 1 by scale No. 2, in which the arc equal 

 to nine divisions of scale 1 is divided into ten divisions. We 

 may from time to time thus change the scales until the 

 instrument has deteriorated in sensibility by one-half, when 

 scale 6 will be used. A little consideration will make obvious 

 that the series of six scales will cover all possible cases of 

 wear between these wide limits. They provide ample and 

 ready means of adjusting the value of the index readings, 



H 



