Heat of Granite and Basalt at High Temperatures. 77 



As the actual observations and method of computation closely 

 resembled the previous ones, there is no need to give another 

 example. The only variation consisted in the insertion of a 

 reversing key in the central circuit. This was needed to 

 obtain a satisfactory determination at 600° C, as about that 

 temperature leakage through the red-hot rock begins to 

 cause trouble. Reversing the central wire current caused an 

 immediate change of about 20 microvolts in the apparent 

 E.M.F.s of the couples, so that readings were taken with the 

 central current flowing both ways, and the means used.. 

 This was quite unnecessary at lower temperatures, as then 

 the insulation was quite satisfactory. 



The results are shown in Table V. and plotted in fig. 7. 



Table V. (Basalt). 



0. 



47°C. 



K. 



W. 



! o- 



K. 



W. 



3935 X 10 ~ 3 



2-95 



1 318° C. 



3-94x10-3 



7-47 



84 



3955 



(512 



; 395 



3-955 



7-90 



110 



4-015 



7-93 



460 



393 



7-68 



200 



4-10 



8-04 



122 



3-795 



8-00 ; 



247 



4095 



8-06 



547 



3 935 



8-20 1 



335 



4-01 



7-97 



! 129 



3-76 



8-29 



390 



3-96 



7-77 



422 



3-925 



8-21 



113 



3-805 



7-52 



600 



3945 



7-90 



203 



391 



7-60 



405 



1 



3-935 



7-80 



Fiff. 



K 



4-2 



x|0-3 







-4 



— ~&~— ^ 













rw-" 



< 







'- ~~"~-~— -*2r 



-> 



„ 



_ 









-J u 



u 









i 













It will be noticed that the first seven results show a slio-ht 

 rise of conductivity up to about 200° C. followed by a fall 

 as the temperature was further raised to 390° 0. This fall 

 was evidently of a permanent nature as shown by the re- 

 maining results. These lie on a lower curve which becomes 

 a straight line above 270° C. From that temperature up to 

 (>00° C, K is independent of 6. The last low temperature 

 determination lies somewhat below the curve, indicating 

 further permanent lowering of K. 



These results resemble the first results obtained with the 

 previous basalt cylinder in showing a slight initial rise of 



)0r e 



