PHYSICS. 475 



data, for the same intervals, are given iu the same table. When cor- 

 rected, the maximum error for the fifteen thermometers is Qo.Sat 200°, 

 0O.5 at 300° and Qo.S at 330°. ( Comptes Bendus, November, xcv, p. 836.) 

 In a third note, Crafts discusses the exactitude of the measurements 

 made with the mercurial thermometer. {Comptes Benclus, November, 

 xcv, p. 910.) 



Siemens has communicated to the Royal Society the results of the ex- 

 periments made with his deep-sea electrical thermometer on board the 

 United States Coast and Geodetic Survey steamer Blake, Commander 

 Bartlett, in the summer of 1881, forwarded to him bj' J. E. Hilgard, the 

 Sn])erintendent. The apparatus was specially ordered for the Govern- 

 ment by Alexander Agassiz, and seems to have proved its reliability. 

 {Nature, June, XXVI, p. 190.) 



Brown has given an extended resume of the various devices brought 

 forward as temperature regulators. {Nature, June, xxvi, p. 111.) 



2. Expansion and Change of state. 



Russner has determined the expansion coefficient of several substances 

 b3' means of the weight thermometer. Sulphur gave from 0° to 10°, 

 .000; 47; 0° to 20°, .000160; Oo to 30°, .000170; to 40°, .< 00178; to 50°, 

 .000183; and to 60°, .000186. Soft rubber of commerce, from 0^ to 10°, 

 .000657 ; 0° to 20°, .000665 ; 0° to 30==^, .000670. Uard rubber has a coeffi- 

 cient between that above given for caoutchouc and that for sulphur. 

 Different specimens varied between .00023 and .00041. ( Carl. Rep., xviii, 

 p. 152; J. Phys., April, II, i, p. 193.) 



Lebedeft" has sought to ascertain whether the apparent contraction 

 of stretched caoutchouc by heat takes place in all directions or only in 

 that in which it is stretched. He finds the latter to be the true hypothe- 

 sis. He has found the cubic expansion coefficient of the rubber to be 

 .000687 when stretched and .000675 when in its natural condition. (J. 

 Soc. Phys. Chim. Busse, xiii, ]). 246; J. Phys., December, II, i, p. 576.) 



Hertz has studied the conditions of the evaporation of liquids, and 

 has measured with great care, by means of a special apparatus, the 

 pressure of the saturated vapor of mercury at different temperatures. 

 The values obtained are lower than those of Regnault. Between 0° and 

 40° the elastic force of mercury vapor varies from 0.00019 to 0.0063™"^ 

 Hence, at ordinary temperatures the elastic force of the vapor of mer- 

 cury is less than one-thousandth of a millimeter. ( Wied. Ann., xvii, 

 pp. 173, 177; J. Phys., November, II, i, p. 512.) 



Cailletet has suggested the use of liquefied gases, especially ethylene, 

 for producing low temperatures. Ethylene liquefies at 10° C. under a 

 pressure of 60 atmospheres, at 8° under 56, at 4P under 50, and at 1° 

 under 45. Its critical point is near -|- 13°. When a thermometer contain- 

 ing CSa was immersed in the liquid it indicated a temperature of — 105°, 

 much lower than that of hyponitrous oxide, which boils at —88°. Spe- 



