Heat of Granite and Basalt at High Temperatures. 61 



estimated by measuring the current and the potential differ- 

 ence between two points at a measured distance apart on it. 

 As the distance between these points was considerably 

 reduced and the diameter of: the platinum wire was increased 

 from 0'3 to 0'4 mm., the P.D.s to be measured never 

 exceeded 1*8 volt, and thus could be directly read on a 

 Tinsley potentiometer. This instrument was used for all 

 the measurements, the " potential " terminals being directly 

 connected to the thermo-couple key, before described, and 

 never disturbed, and the " standard cell " terminals to a 

 specially made switch such that connexion could be made 

 either to the standard cell, or to the points on the central 

 wire whose P.D. was required, or to the ends of a small 

 standard resistance carrying the heating current, thus 

 enabling the latter to be accurately measured. This switch 

 could be turned so as to disconnect the "standard cell" 

 terminals of the potentiometer from all conductors when the 

 E.M.F.s of the couples were being measured. In this way 

 the use of the potentiometer for measuring the energy 

 supplied to the central wire in no way influenced the 

 accuracy of its measurement of the small E.M.F.s of the 

 couples. The original galvanometer was replaced by another 

 giving a deflexion of about 0*5 scale-division per microvolt. 

 This was amply sensitive and much more convenient to work 

 with. For use in measuring the large P.D.s it could be 

 shunted by means of a suitable switch. 



The thermo-couples were, as before, made of platinum 

 and platinum-iridium wires 0*1 mm. in diameter, the junctions 

 being fused. Many of the junctions were formed by lightly 

 twisting the ends together and then passing sparks from a 

 coil to the tips. The wires were then untwisted, and a very 

 small junction was the result. This is a very convenient 

 method of electric welding small junctions, as the heat is 

 developed exactly where it is required. The thermo-electric 

 constants of the wire were obtained from the National 

 Physical Laboratory, but as a larger number of points was 

 found to be required for the accurate plotting of thermo- 

 electric power against temperature, a couple made of the 

 wires was placed in the furnace along with a platinum 

 thermometer and accurately standardized. The figures so 

 obtained agreed perfectly with those furnished by the N.P.L. 



Two patent china cements named " Potmend " and 

 *' Csementium ,J were used in putting the rock cylinders 

 together. The former, which can be made very thick, was 

 found very convenient for filling the grooves, as it was not 

 so hard as " Csementium " and so enabled the wires to be 



