OF ARTS AND SCIENCES. 45 



IxvESTiGATioxs OX LiGHT A^^) IIeat made axd published wholly or in part with 

 Appropriation from the Rumford Fund. 



II. 



CONTRIBUTIONS FROM THE PHYSICAL LABORATORY OF THE 

 MASSACHUSETTS INSTITUTE OF TECHNOLOGY. 



XXL— COMPARISON OF ALCOHOL THERMOMETERS 

 BAUDIN 8208 AND 8209 WITH THE AIR THER- 

 MOMETER AT LOW TEMPERATURES. 



By Axthoxy C. White. 



Communicated by S. TV. Holman, June 10, 1885. 



The results given in this paper were reached in a somewhat careful 

 study of the alcohol thermometer made in the Physical Laboratory of 

 the Institute of Technology during 1881-82. The study, though by 

 no means exhaustive, developed some points of value which are here 

 given. The comparison with the air thermometer agrees substantially 

 with that published by Jolly.* 



The alcohol thermometers used were two centigrade tubes, made 

 by Baudin, Paris, in 1880. They were precisely similar, and of the 

 following dimensions : total length, 50 cm. ; length of bulb, 4.5 cm. ; 

 length of 1° C, 0.28 cm.; divided to 0°.l C. ; range from -|-35° to 

 —90° C. 



The air thermometer was of the Jolly type, but the mercurial 

 columns were read by a cathetometer. Its bulb was of small size, 

 diameter 3.48 cm., to insure a rapid equalization of the temperature 



within and without, and to allow the use of a small bath of liquid 



nitrous oxide. 



The cathetometer was by Staudinger. Its scale was in millimeters 



with vernier reading directly to o'o mm. and by estimation to 0.01 mm. 



The scale errors had been determined by Prof William A. Rogers 



of Harvard University, and corrections were applied. The barometer, 



by James Green, was of the Fortin type, and read to 0.002 inch. 



The scale and instrument errors had been previously determined. 



* Fogg. Ann, Jubclband (1874). 



