NOTE suR l'unit6 de pression. 75 



La nouvelle unite doit s'appliquer a tous les cas de I'elasticite. II 

 convient de ne prendre aucune decision pour la thermometi'ie avant d'avoir 

 approfondi d'une part les simplifications qui r^sulteraient pour la pbys-ique 

 des fluides et notamment la loi des ^tats correspondants de I'emploi d'une 

 seule unite, et, d'autre part, la perturbation qu'introduirait dans la 

 thermometrie et les sciences derivees un changement des bases de I'echelle 

 des temperatures. 



Alloys. — Report of the Committee, consisting of Mr. F. H. Neville 

 {Chairman and Secretary), Mr. C. T. Heycock. and Mr. E. H. 

 Griffiths, appointed to investigate the Nature of Alloys. 



The Committee on alloys beg leave to report that Messrs. Heycock and 

 Neville have been continuing their study of the copper-tin alloys. 



A preliminary statement of the results obtained has been published 

 in the ' Proceedings of the Royal Society,' vol. Ixviii. 1901, pp. 171-178. 

 A fuller account will be presented to the Royal Society shortly ; in the 

 meantime the following summarises their conclusions. 



The work has been directed towards a verification of Roozeboom's 

 theory of solid solutions in its application to the copper-tin alloys. 

 Pyrometric observations have shown that when one of these alloys cools 

 from a high temperature at which it is completely liquid there is often 

 an evolution of heat, not only at the freezing point, but also at one or 

 more temperatures far below that of solidification. This is well seen in 

 the cooling curves published by Sir William Roberts-Austen and Dr. 

 Stansfield some years ago in their reports on alloys. We have found it 

 convenient to repeat some of these cooling curves, which show very well 

 the remarkable nature of these lower halts and the large amount of heat 

 evolved at them. Roberts-Austen and Stansfield have shown in their 

 fourth report on alloys, and more recently in their paper on alloys pub- 

 lished in the ' Proceedings of the Congr^s International de Physique,' that 

 if a continuous line in the concentration temperature diagram be drawn 

 through these lower halts a curve is obtained very similar to a freezing- 

 point curve. We have reproduced this curve so far as our cooling curves 

 enable us to do so, and in the figure the line C'XD'YE' is a copy of this 

 curve. Our cooling curves and the C'E' curve have a certain value as 

 confirming the original ones of Roberts- Austen and Stansfield, but we are 

 not prepared to say that they contain anything new ; in fact our C'E' 

 curve is incomplete. We traced these curves because they were needed 

 for our later work. 



In our figure the upper curve ABCDE is the freezing-point curve — the 

 ' liquidus ' curve, as Roozeboom calls it. The dotted line Ablcde is a 

 rough drawing of the ' solid us ' curve of Roozeboom so far as our experi- 

 ments determine it. This curve is defined by the statement that when 

 the temperature of an alloy falls below the ' solidus ' it sets to a solid 

 mass ; the ' solidus ' might in fact be called the melting point curve. The 

 dotted line IC is a continuation of Roberts- Austen and Stansfield's curve. 

 The numbers at the base of the figure give the atomic percentages of tin 

 contained in the alloys, so that Dl3' on the 20 line corresponds to Cu4Sn, 

 and EE' to CusSn. As will be seen, the figure does not deal with alloys 

 much richer in tin than the latter formula. 



As a microscopic study of the alloys, made in conjunction with a study 

 of the freezing-point curve, has proved that in many cases the structure 

 of the alloys could not possibly have arisen during solidification, but 



