02 Thermal Conductivity of Rocks at High Temperatures. 



than the average for the whole length of wire in the rock 

 cylinder. As it is the latter value thai is used in calcu- 

 lating K, this error tends to make the value of K found too 

 large. 



It is evident that these errors will bo greatly influenced 

 by the rate of heat evolution in the central wire. As in the 

 case of the limestone, no regular change in K is noticeable 

 with change in \V. and the same holds good for granite 

 except when 0> 250° and W<8: it would seem as if the 

 errors are not very important. It must be remembered that 

 the ratio of the length of the cylinder to its radius compares 

 very favourably with the ratio of the diameter to the thickness 

 of the plate used in many guard-ring determinations. 



In experiments at present being conducted on basalt, the 

 dimensions of the furnace have been so modified as to greatly 

 reduce all these possible errors. It is hoped to repeat the 

 determination for granite with this furnace and new cylinders 

 of rock. 



It has recently been found that thermal effects at the 

 potentiometer may introduce small errors, but these generally 

 only amount to a few microvolts, so that their effects would 

 not be of much importance, especially in those determinations 

 in which W and hence A are large. 



The permanent lowering of the conductivity when the rock 

 is heated under atmospheric pressure unfortunately detracts 

 considerably from the value of the results from a geophysical 

 standpoint, as the conductivity of heated rocks beneath the 

 earth's crust may be considerably greater than the figures 

 given above. On the other hand, the conductivity of the 

 deep-seated rock is almost certainly less than the value 

 obtained at first for the unaltered rock at temperatures 

 below 100° C, so in the case of granite similar to the 

 specimen used it would seem probable that for tempe- 

 ratures up to 450° the conductivity must be between 0*0042 

 and 0-0050. 



In conclusion I wish to express my gratitude to Dr. Joly, 

 at whose suggestion the work was carried out, for his kind 

 interest and many valuable suggestions during its progn as 



Physical Laboratory, 



Trinity ( lollege, Dublin, 



April 29, 1912. 



