GEOPHYSICAL LABORATORY. 95 



High Temperature Measurement. — The second investigation was con- 

 cerned with the attainment of means for the more accurate measurement of 

 temperatures, beginning with the low temperatures of every-day Hfe and 

 extending upward to the melting-points of the most refractory metals and 

 oxides. In this investigation, also, it has proved practicable to increase the 

 accuracy of measurements of high temperatures considerably — perhaps ten- 

 fold — and to provide a temperature scale of most trustworthy character in 

 terms of which the temperatures where the various component minerals of 

 the earth are stable can be accurately determined and expressed. The high- 

 temperature portion of this scale was completed a year ago and is already in 

 use; but the lower temperatures (300° to 650° C.) have been redetermined 

 during the present year with especial care in order to meet the more exacting 

 requirements of certain investigations which have been undertaken at these 

 temperatures. The accuracy now attainable in temperature measurements 

 in the vicinity of 500° C. is perhaps ± 0.1° ,* in the region about 1500° C, 

 perhaps ± 2.0°, 



High Pressure Measurement. — Following these two completed investiga- 

 tions, which are of vital importance to all the studies thus far undertaken in 

 this laboratory, it remains to speak of the third : the problem of the accurate 

 measurement of extreme pressures. It is but a few years since all measure- 

 ments of very high pressures were inseparably bound up with such factors 

 as the friction of pistons, the viscosity of the transmitting medium, and other 

 mechanical factors which sometimes affected the final result by as much as 

 30 or 40 per cent. Through the recent efforts of Tammann of Gottingen, of 

 Bridgman of Harvard, and of some others, these uncertain factors have been 

 gradually disappearing, and it is now possible to obtain measurements even 

 of very high pressures with no error greater than that which obtains in the 

 measurement of high temperatures noted above, provided the temperature 

 which prevails in the apparatus is not far above or below the temperature of 

 the surroundings. It is just here that our peculiar difficulties arise, for 

 practically all applications of high pressure in the study of the mineral rela- 

 tions require the simultaneous application of high temperature, and no 

 containing vessel is known which can withstand extreme pressure when hot. 



Investigation along these lines has proceeded but slowly on account of the 

 serious technical difficulties involved in such work, but it is now possible to 

 report that many of these obstacles have at last been overcome, so that a 

 more rapid rate of progress in the future seems assured. Limitations of 

 space prevent more than a mere reference to these obstacles ; the sentences 

 following will serve to indicate some of the points in which progress has 

 been made and to show that the accessible range of pressure and temperature 

 has been extended and the accuracy of measurement improved. 



In the report for 19 11 (Year Book No. 10, p. 95) there was given in 

 abstract an account of work in which temperatures up to 400° and pressures 

 up to 2,000 atmospheres (30,000 pounds per square inch) were employed; 



