628 HENKY A. ROWLAND 



practical measurement to-day give us galvanometers of all kinds and 

 the electro-dynamometer of Weber. By the galvanometer we can meas- 

 ure the quantity of electricity passing at any moment, but by the elec- 

 tro-dynamometer we measure the integral square of the current, a 

 quantity on which the heating of the circuit and the energy expended 

 depend. 



Thus the electro-dynamometer measures the energy from an alternat- 

 ing current dynamo-electric machine as easily as from one giving a con- 

 tinuous current, but to know this energy we must know something else 

 besides the integral square of the current, and this is either the resist- 

 ance of the circuit or the electromotive force. But the measurement of 

 electromotive force depends on a resistance. The question then comes 

 up as to what unit of resistance is the proper one. Here we have to 

 refer to the mathematical theory of the subject, and the great law of the 

 conservation of energy tells us that what is known as the absolute unit 

 of electrical resistance is the proper one for use in this case. Hence 

 the great practical use of determining this unit. The experiments of 

 Kirchhoff, Weber, Kohlrausch, and the British Association found a 

 value from 1 to 3 per cent too large. 



Many years ago I myself experimented on the subject, and obtained 

 a result about 4 per cent too high. Recently Lord Rayleigh has taken 

 up the matter and made a series of experiments of unparalleled accu- 

 racy in this line. The International Commission, determined on by the 

 Electrical Congress in Paris in 1881, met in April of this year at Paris, 

 and has now given us a legal ohm defined as being the resistance of a 

 column of mercury 106 centimetres long and 1 millimetre in section at 

 C. The length best satisfying the experiments is about 106-25, but 

 it was considered best to use the round number. The experiments 

 which I have been making under an appropriation from the Government 

 are now barely completed, but they will probably agree very well with 

 the latter figure. Hence, we can say that we now know this unit of 

 resistance to one part in one thousand, at least. And so we are in a posi- 

 tion to measure the energy of a current to the same degree of accuracy, 

 as far as this quantity is concerned. 



But to measure a current by the tangent galvanometer one requires 

 to know the intensity of the earth's magnetism, a quantity difficult to 

 determine and constantly varying with time and place. The electro- 

 dynamometer, when made with care, is excellent, but a good one is im- 

 mensely expensive. Our methods, then, of current measurement are 

 bad, unless carried out in a completely equipped physical laboratory. 



