GEOPHYSICAL LABORATORY. 97 



(14) Vorlaufiger Bericht iiber das ternare System CaO-AUOa-SiO:. Eine Unter- 



suchung uber die Konstitution der Portlandzement Klinker. E. S. Shepherd 

 und G. A. Rankin; nebst optischen Untersuchungen von Fred. E. Wright. 

 Z. anorg. Chem. 71, 19. 191 1. 



A German translation of the "Preliminary report on the ternary system 

 CaO-AlgOs-SiOo. A study of the constitution of Portland cement clinker" 

 (Jour. Ind. Eng. Chem., 3, i, 191 1). Reviewed under No. i above (p. 92). 



(15) Recent advances in high-temperature gas thermometry. Arthur L. Day. Trans. 



Faraday Soc. London. Vol. vii, 191 1. 



An address before the Faraday Society of London, reviewing the more 

 important features of the recent work with the gas thermometer conducted 

 at this laboratory. This work is briefly described under No. 20 (p. 99). 



(16) Die Untersuchung von Sihkaten. Arthur L. Day. Z. Elektrochem., 17, 609. 191 1. 



An address given at the annual convocation of the Bunsen Gesellschaft, 

 held at Kiel, Germany, reviewing some of the more interesting phases in the 

 laboratory investigation of silicates compared with some of the more familiar 

 fields of physico-chemical research. 



(17) The detection of small heat effects at high temperatures. (Abstract.) Walter P. 



White. Phys. Rev., 32,604. 191 1. 



In calorimetric work at ordinary temperatures, stirring is generally neces- 

 sary to equalize temperatures. Although the need for it is enormously 

 greater in the hot electric furnace, where the surrounding temperature 

 diflferences are much greater, and convection and radiation are far more 

 active, the difficulty of providing for stirring by means of the few materials 

 available at high temperatures is so great that stirring has almost never been 

 tried, and observers have been satisfied to do as well as they could with the 

 uneven temperatures which prevailed in its absence. This has been made 

 easier because the work done has dealt largely with melting-points, where 

 quantitative measurement of heat is not required. Although the results are 

 far less perfect than can be obtained at low temperatures, they have generally 

 proved sufficient for the purpose in view. Some experimenters, however, 

 have applied the melting-point arrangements to the quantitative determina- 

 tion of heat effects, even claiming more than i per cent accuracy. Others, 

 on the contrary, have shown a disposition to reject almost all evidence as to 

 heat quantity based on observations made with the electric furnace. A fair 

 opinion seems to lie between these two views. Determinations in the electric 

 furnace usually show an almost surprising reproducibility, which tends 

 strongly to convey a false impression of accuracy. Large systematic errors, 

 however, are usually present, especially where only one or two thermo- 

 elements are used, since the temperature at any given point affords a very 

 imperfect indication of the average value of the temperature in either fur- 

 nace or charge, while the distribution in both is liable to change both with 

 time and with temperature. It follows, then, that where reproducibility is 

 the main essential, results of high accuracy can often be obtained, but for 

 absolute results the very greatest patience and thoroughness are needed in 

 varying the conditions and otherwise eliminating possible systematic errors. 



In the detection of slight heat effects, high absolute accuracy is not ex- 

 pected. Hence, in this direction relatively great success can be obtained. 



The general principle of the method is simple and fairly familiar. No 

 moving bodies are present, and changes in the temperature of the charge are 

 brought about solely by changing the furnace temperature. The heat effects 



7 — YB 



