On a New Thermometer* for Measuring high Temperatures. 423 



The accompanying diagram is not done to scale, but is merely 

 drawn to exemplify the process. In actual work I use a sheet 

 25 x 25 inches, divided into 100 x 100 squares, each of a quarter 

 of an inch, with every fifth line strong, to assist in counting. 



Royal School of Naval Architecture, 

 April 30, 1868. 



XLIX. On a New Thermometer for measuring Temperatures above 

 the Boiling-point of Mercury. By M. Berthelot*. 



[With a Plate.] 



PROBABLY every chemist and physicist has had occasion to 

 feel the want of a thermometer capable of measuring high 

 temperatures, of portable size, and sufficiently delicate in its in- 

 dications. In organic chemistry especially, and in the course of 

 fractional distillation (an operation of frequent occurrence), this 

 want is very generally experienced. Up to 330° or 350° the 

 mercurial thermometer answers every requirement — although as 

 these high temperatures are approached its indications require 

 correction to the extent of 15° or 20°, since the height of the 

 column outside the distilling apparatus has then to be taken into 

 consideration. But beyond this point we possess no thermometer 

 capable of similar indications. 



In certain cases only are we able to measure elevated tempe- 

 ratures with anything like precision. The air-thermometer, 

 especially with the latest modifications which M. Regnault and 

 M. H. Deville have given to it, certainly allows of such measure- 

 ments ; but the apparatus occupies considerable space, 300 to 

 400 cubic centims. at least — a circumstance which limits its use 

 to the determination of the temperatures of large volumes, and 

 consequently deprives it of all its sensitiveness. On the other 

 hand, the air-thermometer, as it has been hitherto employed, 

 never directly indicates the true temperature ; this is only de- 

 rived from a long calculation in which numerous corrections 

 occur. 



A sensitive thermometer of a size small enough to allow of its 

 use in operations conducted in retorts or flasks, and indicating 

 directly and without correction temperatures beyond 350°, was 

 therefore a desideratum. 



Such an instrument I believe I have succeeded in construct- 

 ing. The thermometer which I am about to describe allows of 

 measurements up to 500°, and may even be employed up to the 

 temperature at which a glass retort commences to fuse. It is 



* Translated from the Annates de Chimieet de Physique, February 18G8, 

 bv Mr. T. E. Thorpe. 



2 F 2 



