166 REPOET — 1892. 



Yard. — The accepted value of the metre, according to Kater, is 



39-3708 inches ; 

 or, according to Clarke, 



39-3704 inches. 



These numbers were deduced from standards of the metric system 

 ■whose value was insufficiently known. But, thanks to modern determi- 

 nations, we have beeu enabled to obtain an accurate reduction of these 

 ancient values, as well as those of Mr. Comstock ; M. Tittmann has thus 

 found 



lm=39-3698 inches. 



According to other measurements by Clarke we should have 



lm=39-3699 inches. 



If we adopt this last value, which seems at present the most probable 

 one, we have 



1 yard =0914404 metre. 

 1 foot =0-304801 „ 

 1 inch =0-025400 „ 

 1 metre=l 093608 yard. 

 1 „ =3 280825 feet. 



Electric Standards. — A great number of researches have been devoted 

 to the measurement of the variation of the specific resistance of mercury 

 with temperature ; the divergences of these results induced me to under- 

 take the redetermination, and to spare no pains to obtain an accurate result. 



I would not trespass on your patience by describing the precautions 

 I have taken, the difficulties encountered in the work, and the reasons 

 which make me leel that this result is sufficiently free from systematic 

 errors. By sixty-four series of measurements made between 0° and 61°, 

 I have arrived at the following mean result : — 



1. Apparent resistance of mercury in hard glass, in terms of the 

 hydrogen thermometer : 



'•t=»-o(1+0-0008809T-(-0000000999T2). 



2. True resistance of mercury (corrected for dilatation of the glass) 

 in terms of the same scale : 



|Ot=Po(1 + 0-0008881T-^0-000001010T2). 



Second Paet. — Propositions. 

 Introduction. 



I wish first to demonstrate by some examples the three following 

 principles : — 



1. There would be danger, from the point of view of precision, in sup- 

 pressing intermediary units which are theoretically unjustifiable, but 

 which can be represented by precise standards. 



2. In certain cases it would be advantageous to reduce the constants 

 to the C.G.S. system more than has been done. 



3. For ordinary physics and industrial application certain approximate 

 definitions are clearly sufficient, while metrologists can, for their own use, 

 apply the necessary corrections to their results. 



