Professor W. J. Sottas — Underground Temperature. 503 



Rock-salt heads the list, and consequently in borings made through 

 this mineral the thermometric gradient should be lower than the 

 average. If, now, we turn to the results given on p. 88 of the 

 British Association Eeport for 1882 we find that the deep Sperenberg 

 boring, which passed chiefly through rock-salt, shows a temperature 

 increase of 1° for 51^ feet, and this result is regarded by the 

 Committee as so remarkably accurate that the effect of quadrupling 

 it when calculating a mean rate is thought worthy of consideration. 

 On p. 84 we read, "The Sperenberg bore, near Berlin, in rock salt, 

 with a depth of 3,492 English feet .... gave an average of 

 1° in 51 "5 feet. This result is entitled to special weight, not only 

 on account of the great depth, but also on account of the powerful 

 means employed to exclude convection." The mean result for all 

 observations given in the same Report is 1° for 64 feet, which was 

 corrected in a later Report to 1° for 60 feet. 



Thus, the gradient of the Sperenberg bore, so far from being below 

 the average, such as the conductivity of rock-salt would have led 

 a believer in the uniform rate of loss of heat to expect, actually rises 

 above it. The average rate at which heat escapes through the earth 

 is given in the Report (1882) as 41-4 gramme degrees annually 

 through each square centimetre of a horizontal section of the earth's 

 substance. There is an error in this number, no doubt typographical ; 

 it should read 51 •4. 



Let us calculate from the data afforded by Sperenberg the average 

 flow of heat through the rock-salt of that district. The gradient of 

 1° in 51^5 feet reduces to 0^0003537 of a degree Centigrade per 

 centimetre. The conductivity of rock-salt, according to the Report, 

 is •0113 X 1^4= ^01582 and 00003537 x -01582 = 55955 X 10"^ 

 which is the flow of heat in gramme degrees per second across one 

 square centimetre, or 55955 X 315 x 10"* = 176'2 gramme degrees 

 per year per square centimetre. In other words, if the influence of 

 conductivity be fairly considered, it leads to the conclusion that the rate 

 of escape of heat at Sperenberg is more than thrice as great as that of 

 the mean (51*4). The data at our disposal in the case of coal-mines 

 do not aj)pear to be sufficient for the purposes of discussion ; all that 

 can be said is, that while the sinkings were made in similar rocks 

 the temperature gradients obtained differ widely among themselves. 

 Without detailed information of the thickness and nature of the 

 various beds passed through in the several cases from which the 

 average is reduced, calculation is impossible. 



A matter of extreme importance has, however, to be mentioned 

 in this connection. In the Report for this year we read (p. 6, 

 separate copy) : " In some condensed reports of Bergrath Kobrich's 

 communication (but not in the full paper as given in 'Gluckauf '), 

 the irregularities are attributed to chemical action in the coal seams, 

 causing in some cases a heating and in others a cooling ; but in the 

 absence of more direct evidence this explanation seems rather 



