March 15, 1918] 



SCIENCE 



267 



memorial to their son, whose death occurred on 

 February 24. 



By the will of Mrs. Charlotte M. Fiske, of 

 Boston, public bequests to the amount of $130,- 

 000 are made. Tuskegee Normal School, 

 Eoanoke College and Bates College receive 

 $5,000 each and Wellesley College receives 

 $10,000 and the residue of the estate. 



Senator W. C. Dennis, president of the 

 Halifa-x Daily Herald, has presented $60,000 

 to Dalhousie University in memory of his son. 

 Captain Eric Dennis, killed at Vimy Ridge. 

 The gift provides that the university shall 

 found a chair of government and political 

 science. 



Dr. Edwin Bissell Holt, assistant professor 

 of psychology at Harvard University has re- 

 signed, his resignation to take place on Sep- 

 tember 1, 1918. 



At the University of Chicago Dr. Harvey 

 B. Lemon, instructor in the department of 

 physics, has been promoted to an assistant 

 professorship; and Dr. A. L. Tatum, professor 

 of pharmacology in the University of South 

 Dakota, has been made an assistant professor 

 in pharmacology and physiology. 



T. J. MuRRY, formerly associate bacteri- 

 ologist of the Virginia Agricultural Experi- 

 ment Station and associate professor of bac- 

 teriology at the Virginia Polytechnic Institute, 

 has been appointed associate professor of bac- 

 teriology at the State College of Washington 

 and bacteriologist of the Washington State 

 Experiment Station. 



DISCUSSION AND CORRESPONDENCE 



THE NOMENCLATURE OF THERMOMETRIC 

 SCALES 



To THE Editor of Science: Present usage 

 in nomenclature of thermometric scales is a 

 cause of indefiniteness and confusion of ideas, 

 and some revision seems called for. Accord- 

 ingly, I hope the statement of the case which 

 follows' will elicit helpful suggestions and tend 

 toward useful results. 



The consensus of scientific opinion and 

 practise is all but universally in favor of the 



1 See also Monthly Weather Beview, Nov., 1917. 



familiar Centigrade scale of temperatures by 

 which the temperature of melting ice and of 

 condensing steam, both from water and under 

 a pressure of one standard atmosphere, are 

 designated 0° and 100° respectively. By gen- 

 eral consent the value of other temperatures 

 than the two points thus fixed by definition 

 are defined by the normal constant volume 

 hydrogen thermometer of the International 

 Bureau of Weights and ileasures as realized 

 by certain mercury in glass thermometers. In 

 recent years the scale of temperatures defined 

 by the varying resistance of pure platinum is 

 also accorded the status of a thermometric 

 standard when its thermal coeflScient as defined 

 by the Callendar equation is evaluated by ob- 

 servations at the melting and boiling iwints 

 of pure water, and at the boiling point of 

 sulphur imder standard conditions defined to 

 be 444.5° or 444.G° C. 



All other thermometric scales that depend 

 on the physical properties of substances may, 

 by definition, be made to coincide at the ice 

 point and the boiling point with the normal 

 scale as above defined, but they will diverge 

 more or less from it and from each other at 

 all other points. 



To obviate the difficulty which arises be- 

 cause thermometers of different types and 

 substances inherently disagree except at the 

 fixed points. Lord Kelvin proposed many years 

 ago that temperatures be defined by refer- 

 ence to certain thermodynamic laws. This 

 course furnishes a scale independent of the 

 nature or properties of any particular sub- 

 stance. The resulting scale has been variously 

 named the absolute, the thermodynamic, and 

 more recently in honor of its author, the Kel- 

 vin scale. The temperature of melting ice 

 by this scale on the centigrade basis is not as 

 yet accurately known, but it is very nearly 

 273.13°, and that of the boiling point 373.13°. 



Occasions arise with increasing frequency 

 in which meteorologists, physicists, and others 

 in dealing with problems of temperature are 

 required to use an absolute scale or an ap- 

 proximation thereto, and to publish tempera- 

 ture data in those units. It is not convenient, 

 and in many cases not necessary, to adhere 



