54 REPORT — 1875. 



Second Report of a Committee, consisting of Prof. A. S. HerscheLj 

 B.A., F.R.A.S., and G. A. Lebour^ F.G.S., on Experiments to 

 determine the Thermal Conductivities of certain Rocks, shoioing 

 especially the Geological Aspects of the Investigation, 



The original object proposed to be effected by the Committee of devising a 

 simple and direct method of determining- approximately the absolute thermal 

 conductivities of solid bodies, and especially of the rocks of most common 

 occurrence in geological strata, has not yet been entirely carried out. But 

 the experiments described in the last Report were repeated under new con- 

 ditions, which enable the Committee to present, with more confidence than 

 in their last Report, a list of absolute conductivities of the rocks there spe- 

 cified, and to add to the list examples of some further important determina- 

 tions. Many uncertainties remained to be removed from the values given in 

 the former list, the most notable of which proceeded from the looseness of 

 the contact obtained between the points of the thermopile and the two faces 

 of the tested plate of rock. The real difference of temperature between the 

 two faces was not measured, but that between the air- currents round the 

 points of the thermopile in the two sheets of velvet which pressed them 

 agaixist the surface of the plate. Thus difference of temperature two or three 

 times as great as those intended to be measured (and as those which were 

 actually observed when in a few cases the points of the thermopile were 

 solidly cemented to the rocks) were constantly recorded. Assuming that the 

 indications were always too great in a certain fixed proportion (determined 

 approximately by the direct experiments made in a few certain cases), the 

 former list of estimated absolute conductivities was compiled from the use of 

 the velvet-covered apparatus. India-rubber faces about two millimetres (or 

 one twelfth of an inch) in thickness are now stretched tightly over the flat 

 faces of the cylindrical cooler and steam-boiler, and are bound down with 

 wire, air being carefully excluded from these junctions by disiilacing it with 

 a little oil. While the top of the boiler is formed of a thick, circular, brass 

 plate a little convex (about one millimetre high in the centre) on the upper 

 side, a tripod wooden stand, with straight legs passing through the lid of the 

 cooler, rests upon the inner side of its base in the sockets of a flat, circular, 

 perforated brass plate with wide lattice openings, and somewhat less in 

 diameter than the cooler -base, between which and the base of the cooler, 

 again, two thicknesses of coarse wire gauze are introduced, to permit free 

 access of the water in the cooler to the inner surface of its tinned-iron base, 

 when the central rod of the agitator, which contains the indicating thermo- 

 meter, is raised and lowered frequently enough to keep the water in a con- 

 stant state of motion. Two twenty-eight-pound weights, suspended from an 

 iron link or cross bar carried by the tripod stand, communicated in this 

 series of experiments a pressure of abotit 3 lbs. per sqiiare inch to the rock- 

 plate placed for examination between the india-rubber-covered faces of the 

 boiler and the cooler ; and it was expected that the exclusion of air from 

 round the points of the thermopile and from the junctions between the faces 

 might be assumed under these conditions : but as india-rubber offers great 

 resistance to the passage of heat, it was found inconvenient to use it of suffi- 

 cient thickness to accomplish this directly by its elasticity* ; and recourse 



* Altliongli thickly jacketed with a close covering of felt and fur, the cooler, containing 

 2^ lbs. of water, was found to lose heat to the atmosphere at the rate of 0°-0025 F. per 

 minute for each degree of excess of its temperature above that of the surrounding air ; 

 and for an excess of 20° F., which was sometimes reached, the loss being 0°05 F., the 



