Sir "Joseph Petavel 193 



by means of microscopes. The yard under test is then 

 compared with the standard, and thus the slightest 

 divergence is revealed. An error of only one-millionth 

 part of an inch can be detected unerringly. 



That surely should be accuracy enough for anyone. 

 But the Laboratory is not yet satisfied. Its workers are 

 now engaged in defining a yard in wave-lengths of light, 

 because any metal, whether steel, platinum, bronze, or 

 anything else, is liable to alter in length during the course 

 of years, and it is important that at no time should the 

 standard yard or metre change by even an infinitesimal 

 part of an inch. For while a fraction of an inch more or 

 less may be of no account to a woman buying cretonne, 

 it is vital to the manufacturer of ball-bearings, each of 

 which must be of identical size if there is not to be 

 friction, or to the manufacturer of pistons for motor-cars, 

 for all these must fit absolutely if there is not to be loss 

 of power. 



It is the same with the measures of weight. These are 

 stored in the Balance Room, where the staff are at work 

 testing weights for scientists, analysts, chemists, and 

 others who need absolutely accurate standards for their 

 work. These weights are tested against the standard 

 weights in scales so sensitive that they have to be insu- 

 lated against changes in temperature. Even the heat of 

 the operator's body might upset their accuracy, and the 

 tests are therefore made at a distance by means of 

 prismatic reflectors. 



Watches, clocks, chronometers, and other time-measur- 

 ing instruments are tested in a similar way against the 

 Laboratory's standard clocks, which are in turn checked 

 three times a day by time-signals from Greenwich, Paris, 

 and Germany. 



Close by is the Physics Department, which tests clini- 

 cal thermometers at the rate of nearly sixty thousand a 

 month. Before a clinical thermometer is passed by the 



N 



