442 G. Barus — Alloys of Platinum. 



netic theory of light furnishes a theoretical basis for the fact 

 that true conductors are exceedingly opaque. The resistance 

 of solid metals, however intensely they may be heated, is 

 found to increase so long as temperature increases. Never- 

 theless the careful experiments which Grovi* made to interpret 

 an erroneous result of Secchi,f prove that solid metals even in 

 extreme states of white heat remain opaque. In the case of 

 liquid metals at extreme white heat the case is not so definitely 

 established ; and the question relative to the ultimate transpar- 

 ency of liquid metals at very high temperatures is an open one.ij: 

 It is in the direction of ultimate transparency that the ob- 

 served continuous increase of resistance with temperature 

 seems definitely to point. 



It is reasonable to infer that the transition from opaque to 

 transparent! will take place in the region of the critical tem- 

 perature. At least such transition must ultimately occur ; and 

 I am led to conjecture that the said transition from opaque 

 to transparent will be accompanied by a change of the values of 

 the electrical temperature coefficient, passing from the negative 

 value which holds for the liquid metal, to the positive value 

 which will probably hold for the gaseous metal, continuously 

 through zero. The fact that conduction in gases is of an 

 electrolytic nature was pro\ied by Yarley,|| who showed that 

 after the polarization of the electrodes is overcome, gases 

 obey Ohm's law. The electric strength of air is known 

 to diminish rapidly as temperature is increased. Working 

 with hot gases carefully insulated and protected from flames, 

 Maxwell^f was unable to obtain conduction either in hot 

 gases like air or in hot metallic vapor like Hg or !Na. 

 At higher temperatures (red heat) the researches of Blondlot,** 

 confirming the observations of E. Eecquerel,ff prove that 

 hot gases are conductors, and that at temperatures sufficiently 

 high y-oVo" volt is enough to set up a current. Hence in 

 their thermal relations also, gases ultimately partake of the 

 nature of an electrolyte, and the occurrence zero value of 

 the temperature coefficient may be reasonably associated with 

 the critical temperature of the metallic liquid, passing con- 

 tinuously from the liquid into the gaseous state. 



*Govi, Comptes Rendus, lxxxv, p. 699, 1877. 



f Secchi, Comptes Rendus, lxiv, p. 778. 1867. 



% W. Ramsay, Chem. News, lv, pp. 104 and 175, 1887; Turner, ibid., p. 163, 

 1887 ; Professor T. Sterry Hunt has given the question some attention. Kundt's 

 recent experiments (Wied. Ann., xxxiv, p. 469, 1888), on the refractive index of 

 metals will doubtless lead to more definite results than the data now in hand. 



§ The jet of liquid hydrogen escaping from Pictet's apparatus appeared steel- 

 blue, and was opaque for a distance of about 12 cm . 



I Var.ey, Proc. Roy. Soc, xix, p. 236, 1871. 



"If Maxwell, Elementary Treatise on Electricity, ed. by Garnett, 1881, §§138, 139. 



** Blondlot, Comptes Rendus, xcii, p. 870, 1881 ; ibid., civ, p. 283, 1887. 



ff E. Becquerel, Comptes Rendus. lv, p. 1097, 1867. 



