ON THE THERMAL CONDUCTIVITIES OF CERTAIN ROCKS. 137 



0405 ; and half the average limit of correction which it may require is 

 7j per cent., making a part of the whole quantity equal to 0029. Sub- 

 tracting this presumptive correction from the whole mean value, the 

 resulting specific heat is - 376, instead of - 405 as obtained directly from 



the table.* The average value of the ratio — for the six specimens, 



0"01947, corrected by addition to it of the same proportional quantity, or 7\ 

 per cent., is - 02088. Sir W. Thomson's determination of the same ratio's 

 value by reduction of the thermometer records, is 0"02311, 11 per cent. 

 greater than the value given by these direct experiments. But the agree- 

 ment is yet very noteworthy if we reflect that (with the single exception 

 of rocksalt, - 0288) no other description of rock in the present list 



exhibits such a high average value of the ratio — as Craigleith sand- 



6 



stone ; quartz itself being a little inferior to it. It owes this property 



partly to its high conductivity, which in the slightly wetted state that the 



process entails (averaging in the actual experiments less than 3 per cent., 



or less than half the quantity of water needed to thoroughly saturate the 



stone) almost attains that of quartz ; and partly to its low heat capacity 



by volume arising from a specific gravity much less than that of quartz, 



which it enjoys in common with other building sandstones. 



Using for the specific heat by volume of Craigleith sandstone a value, 



0"4625, which somewhat exceeds the above found average thermal 



capacity of the rock, Sir W. Thomson's deduction of its absolute thermal 



k 

 conductivity (from the high value of its ratio - already found by his re- 



ductions) is also higher, - 01068, for this second reason, than anything 

 which the Committee has yet met with among rocks of the ordinarily 

 occurring kinds. It must, however, be acknowledged that in its thermal 

 conductivity this building sandstone ranks so high that only compact 

 quartz surpasses it, as the following series of measurements, made during 

 the past year with the apparatus in the excellent state of permanent 

 efficiency which it acquired last year (by the use of German silver and 

 iridio-platinum wires), very plainly indicate. 



As a check upon the exactness of the measures of the thermal 

 capacities the annexed table will also supply useful comparisons with some 

 well-known standard observations of these quantities for a few minerals 



much used for buildings, and "Bed rock," the more general produce of the quarry, better 

 suited for foundations, were also made from specimens obtained for the Committee 

 from the quarry, before Professor Herschel's visit to it, by Mr. K. Irvine, of Granton. 

 Two tenants of the quarry, before Mr. Hunter's succession to it, have occupied Craig- 

 leith-Hill House, and have vacated it since Mr. Johnson's residence there* and the 

 thermometer hole, or well, seems to have been demolished not long after the four or 

 five years' records of its thermometers were ended and discontinued. It took three 

 months to sink it to the full depth of twenty-four feet, with a diameter of three inches 

 at the bottom and six inches at the top ; it was full of water always to a certain height 

 while it was in progress, and was packed, when the thermometers were placed in it 

 .at their proper depths, with sand. A damp or wet state of the rock (and of the in- 

 troduced sand) it may be gathered from these preserved accounts of its construction 

 must have prevailed in the whole or a great part of the stratum of rock through 

 which the bore-hole passed. 



* Had the measurement of the specific heat of each porous plate been repeated 

 with the rock in its perfectly saturated state, it is evident that the necessity of this 

 correction would have been avoided. It is superfluous, apparently, for the sand- 

 stones and less porous rocks, for which one average error-range (for the wet and dry 

 states together), only, is given in the table. 



