350 PHYSICS. 



Eowland lias presented a memoir to the American Academy ou tlie 

 mechanical equivalent of heat, with subsidiary researches on the mercury 

 thermometer as compared with the air thermometer, and on the varia- 

 tion of the specific heat of water. By means of an apparatus contrived 

 for the purpose, the various mercury thermometers to be afterward 

 employed were compared with each other and with the air thermometer. 

 The nature of the glass was fouud to have a sensible influence on the 

 graduation of the mercury thermometer, and the difterences between 

 the mercury and the air thermometers, even between 0° and 100°, are 

 by no means negligible, being some tenths of a degree in the vicinity 

 of 45°. For determining the mechanical equivalent of heat. Joule's 

 method, revolving paddles in water, was employed. A vertical axis 

 carrying the paddles was driven, by a petroleum motor, the vanes them- 

 selves moving in a water calorimeter. The work done by the friction 

 was measured by the thermometer; that expended by the product of 

 the number of rotations into the moment of the couple necessary to 

 prevent the freely suspended calorimeter from turning on its axis. The 

 number of rotations was recorded on the chronograph; and upon the 

 same paper an electric contact recorded the instant when the mercury 

 column reached a given division of the scale. After suitable reductions 

 and corrections, Rowland finds that the mechanical equivalent of heat 

 is a function of the temperature, being 429.8 at 5°, 427.4 at 15°, 426.4 

 at 20°, 425.G at 30°, and 425.8 at 30°. This result the author ascribes to 

 a diminution of the capacity of water for heat, the specific heat diminish- 

 ing as the temperature increases. {Proc. Am. Acad., 1879, p. 75 ; J. Phys., 

 January, 1881, x, 82.) 



2. Expansion and change of state. 



De Lucchi has determined the expausion-coefiicient of sodium from 

 the density obtained in petroleum at various temperatures. The sodium 

 was cast into a cylinder under naphtlia, and before solidification a fine 

 iron wire was inserted in it, by which it was suspended to the balance. 

 The coefficient of expansion of the petroleum oil used having been deter- 

 mined with great care, that of the sodium was obtained by weighing 

 it first in the cold liquid and then in the same liquid at the required 

 temperature. From the data thus given the relative and absolute 

 coefficients were readily calculated. The mean coefficient between 0° 

 and 90° was found to be 0.0002367. The absolute coefficient increases 

 rapidly with the temperature, being 0.00014178 at 0°, 0.00016570 at 20°, 

 0.0001*9580 at 40°, 0.00025160 at 00°, and 0.00036390 at 80°. Near the 

 fusing poiutthe increase is more rapid. [J. Phys., January, 1881, x, 4L.) 



Comstock has called attention to a variation in the length of a zinc 

 bar at the same temperature. The United States Lake Survey pos- 

 sesses a meter made by Eepsold, composed of a bar of steel and one of zinc 

 so arranged as to form a metallic thermometer. It has also a base- 

 measuring apparatus by the same maker, containing cast-iron tubes four 



