COMPARATIVE EXPERIMENTS. 117 



so large a proportion of the motive power as do the other ma- 

 chines, because it uses up a large part of its power in electrical 

 work and in a small space, in consequence of too much having 

 been sacrificed in the production of local actions. By remedying 

 this defect a very good machine might be made of it. 



We regret that a machine of the Siemens type has not been 

 placed at our disposal. Whatever may be its value, our deter- 

 minations would have been of importance as supplying data 

 for a machine so well known and in which the induced arma- 

 ture is so different as regards its principles from those of the 

 other machines, and which, by the manner the wire is wound, 

 theoretically favours the work. 



Thomson and Houston point out that if it is desired to 

 compare the results obtained by them with those of the Eng- 

 lish Commission it is necessary, on account of the different 

 arrangements of the carbons of the regulator adopted in the 

 English experiments, to divide by 2-87 the numbers repre- 

 senting the condensed radiating power. Still the number 

 obtained to represent the luminous power of the Gramme 

 machine would be too great, being 438 candles, whereas they 

 found it to be only 383 candles. This circumstance indicates 

 that we cannot trust entirely to all these figures, for this very 

 number these gentlemen have found to be too high seems to 

 us in France to be much too low. In all these photometric 

 measures there is evidently a sad want of agreement, which 

 shows that all the experiments must be repeated. 



In the meantime we think that we ought to record here 

 the results of experiments made with the Gramme machine 

 by several French engineers and men of science whose names 

 are a sure guarantee for their accuracy. 



We shall begin with Tresca's experiments, carried out at 

 Sautter & Lemonnier's, with two Gramme machines, the large 

 and the small form, which, driven with a velocity of 1,274 

 turns per minute for the large form and 872 turns for the 

 small form, yielded respectively light equal to 1,850 and 300 

 Carcel lamps. I have given in my Expose des applications 



