PHYSICS. 629 



the poles in such maimer that the deflection produced by the coil of 

 the cosine galvanometer shall be opposite to that produced by the cur- 

 rent in the large outer coil, the coil of the cosine galvanometer is in- 

 clined until the compass needle is brought again to zero. In this way 

 the strength of the dynamo current is obtained in terms simply of the cur- 

 rent from the standard Daniell cell, and the method is independent of the 

 strength of the earth's magnetism and of the special field in which the 

 instrument is placed. {Am. J. *%*., March, 1885, III, xxix, 23(5 ; Phil. 

 Mag., May, 1885, V, xix, 396.) 



Cailletet and Bouty have determined the conductivity of the metals 

 at very low temperatures, in baths of methyl chloride, either alone or 

 mixed with carbon dioxide snow, and of liquid ethylene, the temperatures 

 being determined by means of the hydrogen thermometer. They con 

 chule: (1) That from 0^ to — 100° the formula n = r^ (1 + « t), iu which 

 r„ and r^ represent the resistance at 0° and + t°, t the temperature, and 

 a the increase-coefficient, represents the variation of resistance for the 

 pure metals with sufficient exactness ; (2) that if a be determined by 

 the use of a metallic spiral, the temperatures from 0° to — 100° may be 

 measured by the variation of resistance of this spiral with an error of 

 less than 1°; (3) that for each metal a has a special value, which is in gen- 

 eral somewhat above that of the expansion-coefficient of a gas. Hence 

 the application of the formula above given leads to a zero value for the 

 resistance at a temperature somewhat above — 273o. It follows, there- 

 fore, tliat at exceedingly low temperatures the variation of resistance be- 

 (^omes less rapid. The temperatures calculated from the above formula 

 consequently are, for these low temperatures, too near zero. {J. Phj/s., 

 July, 1885, II, IV, 297.) 



Bartoli has examined twenty-three varieties of carbon with reference 

 to their electric resistance. He concludes that to be a conductor the 

 carbon should not contain more than 1-2 per cent, of hydrogen, and should 

 have been submitted to a temperature not below a red heat. He finds 

 t hat an intimate mixture of twenty parts of paraffin and one of graphite, 

 by fusion, however, conducts so well that plates of it may be employed 

 as electrodes ; and yet elementary analysis would show 14-3 per cent, 

 of hydrogen. He believes, therefore, that the conductivity of carbons 

 is due exclusively to the presence in them of finely divided graphite 

 intimately mixed throughout their mass. {II Nuovo Cimento, xv, 203; 

 J. PhijH., December, 1885, II, iv, 563.) 



4. Electric Spark and Electric Light. 



Edlund has contributed further experimental evidence of the i)osition 

 maintained by him, that an absolute vacuum is a good conductor and 

 that the increase of resistance experienced in the ordinary tubes as the 

 exhaustion proceeds is due to the development of a progressively in- 

 creasing counter electro-motive force at the electrodes, a point which he 

 has now rendered piobable. A glass tube, 30«"' long and 16""" in diam- 



