ON THE THERMAL CONDUCTIVITIES OF CERTAIN ROCKS. [)\ 



other clement), which, with the fine gauge (0-4 mm., rolled to l>2 mm.) of 

 the wires used, was not so destructible as iron, while it yielded with German 

 silver a thermoelectric current whose electromotive force was scarcely less 

 considerable than that obtained with a combination of iron and German silver. 

 The series of twenty-four junctions of dissimilar wires contained in the con- 

 tinuous circuit which enclosed the rock, while opposite to each other (above 

 and below the rock-plate) in twelve pairs, were so distributed equally over 

 its area as to indicate by their total action an average difference of tempera- 

 ture between its faces for all the different points of the area of the plate *. 

 The wires, where not used to touch the plate, were secured to a band of thin 

 leather 4 or 5 inches wide, two similar bands of thin silk above and below 

 the rock-plate forming the rest of their support, so that the rock-plate could 

 be placed between the two silk bands in a flexible loop of a twelve-fold coil 

 of wires, the right half of which consisted of German silver and the left of 

 iridio-platinum half-turns of the coil. The latter were cut through, and 

 being joined to twenty-four ends of German-silver wire in a water-bath, 

 which proceeded from as many teeth of a commutator, it was easy (as described 

 by a sketch of the arrangement in last year's Report) to note the actual mean 

 temperature of cither the upper or the lower set of junctions touching the 

 rock-plate, by varying the temperature of the water in the hath until no 

 current passing through the galvanometer indicated that the water in the 

 bath had reached the same temperature as that of the set of junctions above 

 or below the rock-plate with which the junctions in the water-bath had been 

 connected up. 



Several independent proofs having already been obtained that water, with- 

 out convection, possesses a thermal conductivity which is not only high 

 among liquids, but is actually not inferior to that of some solid rocks whose 

 place is low in the conducting- scale, no difficulty was anticipated in making 

 the wire junctions assume identically the existing temperatures of the rock- 

 faces touching them, nor was the bibulous or porous stratum of thin silk upon 

 which they rested, when soaked with water, expected to vitiate the observa- 

 tions by any inequality of temperature in a water-filni of such exceeding 

 thinness, touching the rock, in which the thermopile wires were placed. In 

 order to press them close, smooth sheets of unvulcanizcd india rubber were 

 placed outside of the wet silk ; and the wires being thus effectually squeezed 

 against the rock with a simple luting of pure water (which, under the pressuro 

 of 4 lb. per square inch on every part of the surface, could nowhere well 

 attain half a millimetre in thickness), the equality of their temperature with 

 that of the rock-face contiguous to them might be regarded as assured. In 

 some of the most porous rocks, as chalk and firestone, the water laid on the 

 silk was nearly absorbed by the stone, leaving the silk damp, but steaming ; 

 and as equally steady and satisfactory observations were yet obtained in theso 

 cases when air and water-vapour must to a great extent have replaced the 

 water-film, it deserves a future trial if steam (and it may be even air) in 

 such an extremely thin film as contained the wires in these experiments may 

 not be as effective a medium of heat-conduction with which to surround the 

 wires as water; but, beyond the evidence that air saturated with water- 



* The thickness of the plate was also similarly gauged with steel calipers at several 

 points, so as to obtain the average thickness, the exact value of which as thus obtained 

 was used in all the experiments described in theso Reports to calculate the conductivity of 

 the plate. 



