30 



The mode of using the scales is as follows : — The standard to 

 l>e compared is placed in the left-hand pan and the comparing 

 standard in the right-hand pan. The divisions and fractions 

 of a division which would be pointed out by the index on the 

 ivory scale attached, if the beam were allowed to come to a 

 state of rest, are ascertained by taking the mean of the extreme 

 of the vibrations when the extent does not exceed the limit of 

 the ivory scale, and it is presumed greater accuracy is thus 

 obtained than if the balance came to a state of rest. The 

 standards are then changed from right to left-hand pans, and 

 similar observations are made. The same process is repeated 

 three or six times or more. Half the difference between these 

 two means shows the difference between the two weights in 

 divisions of the index or ivory scale, and this when converted 

 into parts of a grain shows the actual difference in weight. 



There are two gilt gun-metal copies of the Imperial Stan- 

 dard Pound deposited in the Standards Department, No. 31 of 

 1845, which is '00008 grain heavier than the standard, and No. 

 34, which is "000089 grain lighter. As the combined weight of 

 the two standard pounds is thus correct to '000009 grains, their 

 mean weight may be assumed to be equal to the Imperial Stan- 

 dard, and to form a true basis of comparison. These weights 

 have therefore been used for such verifications, the average of 

 their weight being assumed to be absolutely correct. 



It might at first sight be considered that a weight standard 

 once proved correct would remain so practically for ever. But 

 such is not the case, and variations have been proved to have 

 taken place. 



Weights will increase and decrease in weight according to 

 whether they are used or not. If not used they increase by oxida- 

 tion, and if used they decrease by wear. The official standard 

 pound which had been in use since 1825, when compared in 1866 

 with Captain Kater's primary brass pound, No. 1, which had 

 never been used, and of which it is a copy, was found in the mean 

 of five comparisons to be "90177 grain lighter. This has been 

 assumed to be the loss of weight due to wear and tear in 40 

 years. I consider we must also assume that so careful an ob- 

 server as Captain Kater could not have overlooked so great a 

 difference. As an instance of increasing in weight, Captain 

 Kater's pound, No. 5, was compared in 1824, and again in 1844, 

 and in these 20 years it had gained "0089 grains by oxidation, 

 and upon comparing again in 1867 it had gained "01698 grains 

 in the next 23 years. The accuracy of this has been questioned, 

 but there can be no doubt of an increase in weight. 



An illustration of different forms of one-pound weights, half 

 full size, is given (Plate XI.) to give an idea of their specific 

 gravities. 



