140 



REPORT — 1881. 



Peclet, ib. (3) 2, 107 (1841).— The well-known method. Eesult for Pb k = 3'84 

 (1 kilo. H,0 1° C. 1 sq. m. 1 mm. 1 sec). Hence, from the results of Despretz, 

 Au = 21-28', Pt = 20-9.5, Ag = 20-71, Cu= 19-11, Fe = 7-95, Zn=7-74, Marble = -48, 

 Porcelain = -24, Fireclay = -23. 



8enarmont de, ib. (3) 21, 457, and 22, 179 (1848) Crystals. Covered with wax, 



and heated either by spot of sunlight, by voltaic ignition of Pt wire bored 

 through, by stream of hot air through silver tube, or by conduction in silver rod. 

 Conductivity varies much as optic elasticity does. In cubic system, surface 

 is a sphere ; in square prismatic or rhombic, ellipsoid of revolution, with unequal 

 axis coinciding with axis of symmetry ; in rectangular prismatic, ellipsoid axes 

 coinciding with axes of crystal ; in monoclinic, ellipsoid with one axis in axis of 

 S3Tnmetry, others not predicable ; in anorthic, no axis predicable. 



Senarmont de, ib. 23, 257. Stress and conductivity. Glass, porcelain, and flint 

 compressed, and the old method used. Axis in direction of compression always 

 shortened, and vice versa. 



Wiedemann and Franz, ib. 41, 107 (1854). Metals. The method of Langberg, 

 using thermopile with mechanical means of ensuring contact,, and calibrating the 

 indications by direct comparison with thermometer. Kesults in air diifer slightly 

 from those in vacvo. 



Pt 8-4 

 Ag 100-0 . Brass 23-1 . Fe 11-9 . Pd 6-3 

 Cu 73-6 . „ 24-1 . Steel 11-6 . Kose 2-8 



Au 53-2 . Sn 14-5 , Pb 8-5 . Bi 1-8 



AViedemann, ib. 45, 377 (1855). — Same method applied to junction of two metals. 

 Appeared at first that there was a sudden fall at the junction when better con- 

 ductor was hotter, but none when bad conductor was hotter. This, however, 

 found to be due to rate of commimication of heat to thermo-pair, and when 

 thermo-pair was immersed in Hg in each bar no difference of temperature was 

 found at the junction. 



Gouillaud, ib, (3) 48, 47 (1856).— Metals (Fe, Zn, Pb). Despretz' method, only 

 using longer bars. Verifies experimentally the formulae y = T6~»^, and 

 y = A€" + B6"" between the limits for which Newton's law holds. Verifies also 

 the law of thicknesses as Despretz didfor liquids. Deduces and verifies that the 

 constant A in expression — y = A (e"' — €"*') -t-Te~"' — varies directly as the ex- 

 cess of temperature of the source, and shows tliat without sensible error the 

 formula may be put in the form (2E being the thickness of the bar and I its 



length), y = ^~^^ Tt-'»'(6"-t-") +T6-". 



Wiedemann, ib. (3) 58, 126 (1860). — Alloys. Calvert and Johnson's paper appears 

 to show that with alloys the thermal and electric conductivities do not agree 

 as with metals. According to these results they do. Thermal conductivity 

 determined by W. and F.'s method, electrical by Wheatstone's Bridge comparison 

 with zinc wire. 



40 3-2 



With Cu Zn alloys, conductivity is nearly same as that of Zn. With Sn Bi 

 alloys, conductivity is about the mean of those of constituents. 



Neumann, ib. (3) 66, 183 (1862). — Variable state observed. For metals, bars pro- 

 vided with tbermo-pairs near either end heated to stationary state, then sum and 



