602 Mr. J. Viriamu Jones [May 24, 



We run the risk, therefore, of having a varying unit, and the prime 

 requisite of a unit is constancy. 



In England the Government has taken this risk, and committed 

 itself — at least for the present — to this method of specifying the unit 

 of electrical resistance. 



The denomination of the standard of electrical resistance contained 

 in the recent Order in Council regarding Standards for Electrical 

 Measurements is as follows : — 



" A standard of electrical resistance denominated one ohm, being 

 the resistance between the copper terminals of the instrument marked 

 ' Board of Trade Ohm Standard Verified, 1894 ' to the passage of an 

 unvarying electrical current, when the coil of insulated wire forming 

 part of the aforesaid instrument and connected to the aforesaid 

 terminals is in all parts at a temperature of 15* 4° C." 



This is a plain adoption of an arbitrary standard. The resistance 

 of a certain piece of wire wound into a coil is made the legal unit, and 

 if this resistance varies our legal unit varies with it. 



It is true that in constructing the standard, efforts have been made 

 to ensure that its resistance shall be equal to 1000 million times the 

 absolute unit of which I shall presently speak — that the standard was 

 in construction based upon the ohm ; but there is no provision in the 

 Order in Council for revision if the standard varies, nor indeed has 

 the Government any means at its disposal of directly measuring its 

 standard in absolute measure. 



I seem to be finding fault. That is not my intention. I am only 

 advocating progress. The time is ripe for the adoption of the 

 absolute unit, not merely nominally but really, for the frank acceptance 

 of the absolute unit itself as the ultimate legal standard. 



Now, what is an absolute unit ? The following considerations 

 affecting physical measurements generally will, I hope, serve to make 

 this clear. 



In order to specify the magnitude of any physical quantity, we 

 bring it into relation with a standard magnitude of the same kind, 

 called the unit, and say that it is so many times or such and such a 

 fraction of this unit. Thus we speak of a length of 6 centimetres, 

 a mass of 50 grammes, a time of 20 seconds, an electric current of 

 10 amperes, an electrical resistance of 30 ohms, and so on. 



For the purpose of this specification any convenient magnitude 

 of the physical quantity to be measured may be taken as the unit. 



But science has to deal with many kinds of quantities, and when 

 we consider the various units as constituting a system, our arbitrary 

 choice is limited by two conditions of fundamental importance : — 



1. The units must be chosen so as to simplify as far as may be 

 the statement of the quantitative relations existing between various 

 kinds of quantities. 



Physical science, in so far as it is quantitative — may we not say in 

 so far as it is perfected ? — tells us of relations of interdependence 

 between many different kinds of quantities. Thus, in geometry we 



