138 REfOST— 1881. 



variations at the place of the soil's dryness or Wetness during the same 

 time. The high valae of the stratum-conductivity, found by Quetelet, in 

 two years at Brussels, for earth of the deep-sunk thermometer bed, estab- 

 lished and observed there in 1834 and 1835 (0-0120), may, perhaps, for 

 the same reason, be in some measure accidental ; but no errors of the 

 same kind can be easily supposed to affect the widely different value 

 (0-0057) which Angstrom's reduction in 1851, of seven years' records of 

 the deep-sunk thermometers, at Upsala, gave for the uniform bed of 

 sandy clay in which those thermometers are placed. 



The surface-stratum conductivity given by Poisson's reduction, in his 

 ' Theorie Mathematique de la Chaleur,' of Arago's unpublished e%rth- 

 thermometer observations in Paris (0-0083), and that assigned by Ang- 

 strom's discussion of Caldecott's observations for three years of deep-sunk 

 thermometers in ' laterite,' at Trevandrum (0-0074), are both of average 

 or ordinary magnitudes. Two values, however, obtained subsequently 

 by Angstrom, apparently from earth-thermometer records at Upsala, 

 for wet clay and argillaceous sand (0-0047 and 0*0045), are again re- 

 jnarkably low quantities compared with the well-established conductivities 

 in terms of heat-capacity by volume (00205 or 0231) found by Forbes 

 and Thomson to belong to Craigleith sandstone. 



Of the existence, thei-efore, in different rocks and soils, of great 

 diversity in their powers of transmitting heat-waves, or periodical and 

 other fluctuations of temperature, the hitherto conducted observations of 

 earth-thermometers have given abundant proofs and illustrations. It 

 redounds, indeed, especially to the well-earned scientific prestige of the 

 British Association that this result has been arrived at by one of its 

 earliest initiatives and promotions of scientific objects in no small degree ; 

 for by its timely establishment and careful choice of the sites of the 

 Edinburgh rock- and earth-thermometers, a thermal property of the pure 

 Craigleith coal-measure sandstone has been revealed, to which no other 

 earth-thermometer records and reductions yet made have hitherto pre- 

 sented any parallel. The geothermal conductivity of this kind of sand- 

 stone is nearly twice as great as that of any other earth-stratum hitherto 

 tested with thermometers, and five or six times as great as that of moist 

 clay and argillaceous earth tested by the same mode of observation. 



An error of reduction which affected all the Committee's absolute 

 conductivity-determinations, was noticed after the production in the 

 volume of these Reports for the year 1878, of the last general table 

 of the Committee's observations. The necessary correction which it 

 entailed upon the list was notified in the following report (in the volume 

 of these Reports for the year 1879), and corrected values of several of 

 the results before obtained were then communicated. The required 

 corrections have now been applied to the whole of the measurements 

 contained in the last general table, and the numerous trials made in 

 several cases of rocks of the same class and geological description have 

 been incorporated, owing to the close similarity which in general subsists 

 between them, into single average results for the several distinct classes 

 or geological kinds of rock. The comparison which can thus be formed 

 in the accompanying general list between these measurements and those 

 obtained by other methods and researches is in general very satisfactory, 

 and exhibits in many cases corroborations of very excellent accordance. 



The confirmations desired by Poisson of the applications of Fourier's 

 theory to the problem of heat and temperature- distribution and changes 



