340 PROFESSOR W. C. ROBERTS-AUSTEN ON MECHANICAL PROPERTIES 



copper into the worst conceivable," and by the observation of Mr. PREECE,* that " a sub- 

 marine cable made of the copper of to-day," the necessity for employing pure metal being 

 recognised, "will carry twice the number of messages that a similar cable of copper would 

 in 1858," when less importance was attached to the presence of foreign matter in the 

 copper. It may be well to refer to a but little known case in which the change in the 

 structure of a metal produced by the presence of a minute quantity of foreign matter 

 becomes at once evident by comparing the fractured surfaces of the pure and impure 

 masses. Bismuth, when pure, has a fracture which shows large brilliant mirror-like 

 crystalline planes ; but, if only the 10 1 Q th part of tellurium be present, the fracture is, 

 as a specimen submitted to the Society showed, entirely different, being minutely 

 crystalline and lighter in colour than pure bismuth. 



The mode of action of these small quantities of impurity is still very obscure, but 

 it should be remembered that Professor W. SPRING, of Lie"ge, has recently given 

 evidence t in favour of the view that molecular polymerization may take place even 

 in a solidified alloy, and MATTHIESSEN,J in a classical series of researches on the 

 electrical resistance of alloys, communicated to this Society nearly thirty years since, 

 was led to the view that in many cases the constituent metals of alloys exist in the 

 form of allo tropic modifications, the quantities of the metal producing a rapid decre- 

 ment in conductivity being too small to enable the effect to be explained by attributing 

 it to the formation of chemical compounds. 



In the present paper, attention is directed to the way in which the tenacity and 

 extensibility of metals may be affected by small quantities of metals and metalloids, 

 with the view of showing that the relations between these small quantities of the 

 elements and the masses of metal in which they are hidden are under the control of 

 the Law of Periodicity, which, as originally expressed, states that " the properties of 

 the elements are a periodic function of their atomic weights." CARNELLEY has set 

 forth at some length the reasons for supplementing the law as follows : " The pro- 

 perties of the compounds of the elements are a periodic function of the atomic weights 

 of their constituent elements " ; and the question arises, may the law be so extended 

 as to govern the relations between the constituent metals of alloys, in which, as is well 

 known, the atomic proportions are often far from simple. 



The influence of a small quantity of one metal on another is so marked that it 

 appeared well to approach the consideration of the problem by investigating the 

 nature of the change so effected in the mechanical properties of metals. Gold was 

 the metal selected as a basis for the experiments, mainly because it can be more 

 readily brought to a high degree of purity than any other metal : the accuracy of the 

 results of the experiments are not likely to be disturbed by the oxidation of the gold 



* ' Instit. Civil Engineers Trans.,' vol. 75, part 1, 1883. 



t ' Bull, de 1'Acad. Roy. de Belgique,' voL 11, 1886. 



J ' Phil. Trans.,' vol. 150, I860, p. 85 ; and vol. 154, 1864, p. 167. 



' Phil. Mag.,' vol. 8, 1879, p. 368. 



