itH M. RegnauU on the Measurement of Temperatures 



are the same as those used in my calorimetric experiments: 

 each centigrade degree is divided into eighteen divisions, so 

 that the differences of temperature of the two solderings can 

 be measured with extreme accuracy. 



Lastly, a very sensitive differential galvanometer completes 

 the apparatus. This galvanometer has a system of two little 

 compensated needles, to which is fixed a long hollow glass 

 rod drawn out very fine. The extremity of this rod, which 

 is blackened, moves on a divided quadrant 15 centimetres in 

 diameter; it is observed by means of a telescope. The qua- 

 drant is divided into quarters of a degree ; and it is very easy 

 to appreciate deviations of ^th, and even of y^^jth of a degree. 

 The sensitiveness of the galvanometer is such, that a differ- 

 ence of temperature of 1° in the two solderings of bismuth 

 and antimony produces a deflexion of the needle of 17°. 



The galvanometer is connected with the iron and platinum 

 circuit by one of its wires, and with the bismuth and antimony 

 circuit by its second wire. 



With this arrangement, the junction of the iron and pla- 

 tinum, E, being maintained at a constant temperature ^, if 

 the junction F be raised to a temperature T' measured on 

 the thermometer of the oil-bath, a current will result which 

 will deflect the needle of the galvanometer; but by suitably 

 raising the temperature of one of the solderings of the bismuth 

 and antimony element, a second current, the inverse of the 

 first, will be obtained, by means of which that may be neutral- 

 ized and the needle of the galvanometer restored to 0. The 

 temperatures 5 and 6', which the two thermometers T and T' 

 indicate at the moment of neutralization, are noted down. 



Thus a difference of temperature, T'— /, between the two" 

 solderings of the iron and platinum, produces a current which 

 is neutralized on the galvanometer by the current developed 

 in the bismuth and antimony element by a difference of tem- 

 perature fl'— fi. This difference of temperature, 6' — fl, is more- 

 over much smaller than T' — /, because the electromotive force 

 of the bismuth and antimony element is incomparably greater 

 than that of the iron and platinum element, for with T'—t 

 = 1 00 we have 6' — 9 = 6°-5. 



If the oil-bath be raised to the temperature T", 6' must be 

 raised to 0" in order to maintain the needle of the galvano- 

 meter at 0°. By continuing in the same manner, we obtain a, 

 series of temperatures T' — ;^, T"— /, V" — t, &c., which will 

 produce on the iron and platinum element currents which 

 will hold 171 equilibria on the galvanometer the currents 

 produced on the bismuth and antimony element by the differ-t 

 ences of temperature 9'— 5, 6"— S, P— 5, &c. If, consequently. 



