52 THE PRINCIPLES OF SCIENCE. 



measure electricity as the variant. That one of the two ] 

 measured quantities which is an antecedent condition of 

 the other will be the variable. 



It will always be convenient to have the variable en- 

 tirely under our command. Experiments may indeed be ] 

 made with accuracy, provided we can exactly measure the 

 variable at the moment when the quantity of the effect is] 

 determined by it. But if we have to trust to the action ; 

 of some capricious and very uncertain force, there may bel 

 great difficulty in making exact measurements, and those 

 results may not be disposed over the whole range of 

 quantity in a convenient manner. It is one prime object 

 of the experimenter, therefore, to obtain a regular and 

 governable supply of the cause or force which he is in- 

 vestigating. To determine correctly the efficiency of wind- 

 mills, when the natural winds were constantly varying in 

 force, would be exceedingly difficult. Smeaton, therefore, 

 in his experiments on the subject, created a uniform arti- 

 ficial wind of the required force by moving his models 

 against the air on the extremity of a revolving arm b . 

 The velocity of the wind could thus be rendered greater 

 or less, it could be maintained uniform for any length of 

 time, and its amount could be exactly ascertained. In 

 determining the laws of the chemical action of light it 

 would be out of the question to employ the rays of the 

 sun, which vary in intensity with the clearness of the 

 atmosphere, and with every passing cloud. One great 

 source of difficulty in photometry and the experimental 

 investigation of the chemical action of light consists in 

 obtaining a perfectly uniform and governable source of 

 light rays c . 



b ' Philosophical Transactions,' vol. li. p. 1 38 ; abridgment, vol. xi. p. 355 

 c See Bunsen and Roscoe's ' Researches,' in ' Philosophical Transactions 



(1859), v l- cxlix. p. 880, &c., where they describe a constant flame o: 



carbon monoxide gas. 



