DO 



Proceedings of the British Association* 



and the result of two years' reduetions was presented to the Section.* 

 The object was to ascertain the conducting power for heat of different 

 soils, and the measure of the sun's influence at different depths under 

 similar external circumstances. The stations and soils were, 



Observatory. Experimental Garden. 



Craigleith. 



Trap Tufa. 



Pure loose sand. Compact coal-formation sandstone. 



At each station four thermometers were sunk to the depths of 3, 6, 

 12, and 24 French feet respectively, the tabes of each being carried above 

 the surface, so as to be conveniently exposed side by side. The read- 

 ings were made every week, and corrected for the temperature of the 

 stem and scale, and the results were projected in the form of curves. 

 from which the following deductions have been made : — 









I 



. — Annual 



Rep< 



JRT. 















3 Feet (French). 



6 Feet. 



12 Feet. 



24 Feet. 





p. 

 H 



■6 ] -o oj 

 «> J co tS 



p. 

 d 

 u 

 H 



3 



CO 



co % 



P. 

 M 



H 



d 



IS 

 CO 



13 Q> 



§ 2 



CO % 



09 



H 



■d 



a 



CO 



CO to 



1837. Fahr. 

 Cent. 



1838. Fahr. 

 Cent. 



18.95 

 10.53 



17.7 

 9.83 



] 9.65 

 11.23 



20.33 

 11.30 



17.25 

 9.58 



18.52 

 10.29 



11.9 

 6.61 



11.2 



6.22 



14.95 

 8.30 



14.57 

 8.10 



13.9 

 7.72 



14.25 

 7.91 



5.5 

 3.05 



5.05 

 2.80 



7.55 

 4.19 



7.1 



3.94 



9.4 

 5.22 



9.3 

 5.16 



1.45 

 0.80 



1.25 

 0.70 



2.1 

 1.16 



1.88 

 1.05 



4.1 



2.28 



3.83 

 2.13 



These numbers involve the data for computing the Conductivity of 

 these several strata; for the range in each case is found (as theory 

 indicates) to diminish in geometrical progression, as the depths increase 

 arithmetically, and the common ratio of the progression depends on 



the value of / Specific heat , which is the value of B in the follow- 

 */ Conductivity 



ing formula A p = A-f-Bp. 



Where & p is the range, and p the depth from the surface. To obtain 

 the value of B, the above ranges were projected, and logarithmic curves 

 drawn through the points, so as to satisfy approximately the observa- 

 tions at each station; the result is shown for the year 1838 in the sub- 

 joined figure, from which it will be seen that the experiments are 

 perfectly consistent with one another, although (as at first sight might 

 be expected) the amount of the range does not follow the same order of 

 magnitude in the three soils at different depths. 



The values of B, thus deduced, are — 







Trap. 



Sand. 



Sandstone 



In 1837 —.0545 



- .0440 



— .0316 



In 1838 —.0641 



— .0517 



— .0345 



* Some account of the first year's experiments has already been presented to the 

 Royal Society of Edinburgh, and printed in their Proceedings. 



