OCTOBKR 31, I 901] 



NA TURE 



667 



suhtlivi-sion of subjects. They wanted a chair of geology and 

 much more subdivision in the dift'erent branches of chemistry, 

 and, first and foremost, a chair of the German language and 

 literature. More teaching power was required and more space 

 in which that power might exercise itself. 



The installation of Lord Balfour of Burleigh as Chancellor of 

 the University of St. Andrews, in succession to the late Duke of 

 Argyll, was made the occasion of a series of brilliant functions 

 in the ancient city last week. The address delivered by Lord 

 Balfour after his installation dealt with several important aspects 

 of higher education, and his statement of the relationship be- 

 tween national progress and scientific research should be of value 

 in .showing that the work done in a progressive University is 

 technical training of the highest kind. Referring to University 

 studies, Lord Balfour remarked, " Besides the broad general 

 treatment of any scientific course, the University should be 

 enabled, as a sequel to that course, to specialise in the more 

 advanced stages of scientific training and to encourage original 

 reseaich on particular lines. For this purpose the University 

 must have full equipment and must be furnished with teachers 

 of special attainments, who will direct and guide that original 

 research. The students will pursue the subject as a University 

 study, and with the view of enlarging and advancing the know- 

 ledge of their special subject. In this way only can a real 

 advance in scientific knowledge be made ; and from the students 

 who pursue these courses — generally, I would say, post-graduate 

 courses — we must look for the enlargement of scientific know- 

 ledge, and amongst them or as the products of their efforts we 

 may find pioneers in the application of truly scientific method 

 to our manufactures. A modern University must deal with the 

 principles which lie at the root of our commercial relations, and 

 upon which the development of manufactures must rest, just as 

 much as it does with the principles of philology and mental 

 philosophy. Our commercial, no less than our educational 

 position, must be supported by a thorough training, by careful 

 attention to principles and by imparting to young men who are 

 to pursue commercial pursuits the power of grasping the wider 

 aspects of ihe questions with which they will have to deal, and 

 by taking care that while they obtain a training fitted to be of 

 practical value to them in their after life, that training shall be 

 such as will really awaken their intelligence and enable them to 

 cultivate the inestimable qualities of judgment, of foresight and 

 of enterprise." Upon this subject scientific and practical men 

 are in agreement, and the demand will be met, as Lord Balfour 

 remarked, not by curtailing the work of the Universities, or by 

 lowering in any way their standard of scholarship or of pure 

 science, but by enlarging their borders and extending their 

 influence. 



SOCIETIES AND ACADEMIES. 



LO.NDON. 



Physical Society, October 25.— Prof. S. P. Thompson, 

 president, in the chair. — A paper on the variation with tem- 

 perature of the thermoelectromotive force and of the electric 

 resistance of nickel, iron and copper between the temperatures 

 of - 200° and + 1050° was read by Mr. E. P. Harrison. In 

 this paper the changes with temperature of the thermoelectro- 

 motive force and the resistance of nickel and iron are traced over 

 a wide range and the singularities present in the curves repre- 

 senting these changes are investigated. In all experiments the 

 same specimens of metal were used. Previous work on this 

 subject has been performed by Tait, Fleming and Dewar, 

 Holborn and Day, and Stansfield. In the author's experiments 

 on E. M.F. an ordinary potentiometer method was used, the 

 potential difference due to the thermocouple being balanced 

 against a portion of that due to two accumulators. Before 

 each reading a standard cadmium cell was balanced on a definite 

 resistance in the accumulator circiiit. Readings of E. M.F. of 

 copper-nickel couples were accurate to i '8 microvolts, while 

 those of coppar-iron couples were accurate to less than one 

 microvolt at moderate temperatures. The heating arrangement 

 was designed to give a uniform temperature which was measured 

 by a platinum thermometer and recorded automatically by 

 Callendar's recorder. The cold junctions were placed in a large 

 test-tube full of water, the test-tube being placed in a larger 

 vessel also containing water. The temperature of the cold 

 junctions varied with that of the room, and all observations 

 were reduced to cold junction 0° C. Finally, in each case, 



NO. 1670, VOL. 64] 



observations were taken by placing the junctions in liquid air 

 with the platinum thermometer beside them. To prevent oxi- 

 dation of the metals forming the junctions at temperatures 

 above 500° it was necessary to exhaust the porcelain tubes 

 which contained them. The curves for variation of E.M.F. 

 with temperature of copper-nickel and copper-iron couples are 

 roughly a straight line and a parabola respectively. The differ- 

 ences between the actual curves and a selected straight line in 

 the former case and a parabola in the latter cas: have been 

 plotted against temperature. Thesedifferencecurves.show that the 

 maximum variations occur, in the case of copper-iron, at 70% 230°, 

 and 370". The temperature of inversion (cold junction 0° C.) is 

 536° C. and the neutral point is 262° C. In tha case of copper- 

 nickel, maximum variations occur at 70° and 340°, and 

 there appears to be a small hysteresis effect at the 

 latter point. The temperature of inversion does not occur 

 within the limits of the experiments, and there is no neutral 

 point. The E.M.F. curve for a nickel-iron couple up to 700° 

 has been obtained from the two previous experimental curves 

 by addition. Above this temperature direct observations have 

 been taken. This curve is nearly linear up 10 900°, at which 

 point a decrease in E.M.F. occurs. Curves of thermoelectric 

 power have been derived from the E.M.F. curves by drawing 

 tangents, and these show that a considerable range of the copper- 

 iron curve can be represented by straight lines, but that the 

 remainder is approximately parabolic. The copper-nickel power 

 curve can be represented by bits of straight lines. The Peltier- 

 coefficient variation curve for iron-copper is at first parabolic and 

 can then be made up of straight lines ; for copper-nickel it can 

 be made up of bits of parabolas. Considerable difficulty was 

 experienced at high temperatures in getting concordant results 

 owing to chemical changes and other effects. The experiments 

 were therefore carried out under different conditions, and the 

 results are discussed in the paper. In the resistance experi- 

 ments a potentiometer method was employed, a manganin resist- 

 ance coil immersed in an oil bath being used as a standard. 

 The resistance of nickel increases with temperature almost para- 

 bolically up to 370°, when a change of slope occurs, and the 

 resistance increases much less rapidly and almost linearly up to 

 1050^. In the case of iron, the resistance curve does not change 

 its parabolic form till nearly 800°, when it becomes linear and 

 remains so up to 1050'. The author concludes from his paper 

 that the thermoelectric change in nickel-copper coincides ap- 

 proximately with the resistance change, but that no thermo- 

 electric peculiarity exists for iron-copper at the temperature cf 

 the iron resistance change. Mr. A. Campbell said that with 

 purer iron the change in thermoelectric properties might corre- 

 spond with the change in resistance. Dr. Knott had performed 

 experiments on nickel in 18S6 and got results similar to those 

 of the author. His results with thick wires were different to 

 those with thin, probably because he did not exclude air and 

 prevent oxidation. Mr. Campbell said that he had himself 

 made experiments upon two samples of nickel differing in 

 resistivity, and although their temperature coefficients were also 

 different, the change in slope of the curve connecting resistance 

 and temperature occurred at practically the same temperature 

 in both specimens. Their thermoelectric powers were identical 

 up to 300°, but above they differed slightly. Dr. D. K. 

 Morris pointed out that the thermoelectric force, the 

 resistance and the magnetic properties should be observed 

 at the same time. In taking a thermoelectromotive force there 

 must be a temperature gradient, and in the interesting parts of 

 the curves differences of magnetic properties may arise and pro- 

 duce discrepancies. He drew attention to the caution which 

 must be exercised in differentiating by drawing tangents 

 except when the curves are smooth. Dr. Morris said the con- 

 nection between resistance and magnetic qualities was interest- 

 ing. The temperature coefficient of resistance of a magnetic 

 body rises with temperature so long as the body is magnetic, 

 but reverses when the body becomes non-magnetic. He asked 

 for information on the subject. Prof. H. L. Callendar said he 

 had followed the research with interest, and referred to the ex- 

 perimental difficulties, especially at high temperatures. He 

 should like to have said something in reply to Dr. Morris, but 

 he was afraid the subject was a large one and might well be 

 discussed at some future meeting. There were several points 

 to clear up, and the fact that the curves described cannot be 

 represented by straight lines or parabolas showed that the 

 subject was beyond the range of a simple theory. The chairman 

 suggested that it might be well to re-examine more carefully 



