60 Mr. 0. Heaviside on the 



The subject was partially investigated by OErsted in 1828*; 

 and as its practical importance was further indicated by Gay- 

 Lussac in 1830f, I made a series of experiments in order to 

 ascertain how far the more delicate appliances in use at the 

 present day could be made available. The results, however, 

 were not entirely satisfactory. 



Prof. Hughes's Induction -balance rendered it possible to 

 resume the research on a new basis. It is only necessary to 

 glance at such a curve as that of the Gold-Silver series No. II. 

 to be satisfied of the probability that certain parts of it, at least, 

 would indicate minute differences of standard. I would there- 

 fore direct special attention to the series of alloys which lie 

 between pure silver and silver alloyed with 5 per cent, of 

 gold. These are shown in a separate curve, where the scale 

 of percentages is more extended. Such alloys as Nos. 2 to 6 

 are known to refiners as dore ; and No. 2 contains less 

 than 2 grains of gold to the pound troy, a quantity which 

 could not be extracted with profit by the ordinary operation of 

 " parting." Small as the amount of precious metal is, its 

 presence is clearly indicated on the induction curve, as are 

 also the larger amounts of gold contained in Nos. 3-5. 



Experiments are in progress on other series which promise 

 to afford trustworthy indications ; but of course the establish- 

 ment of a method of verifying the composition of alloys of 

 the precious metals must in part depend on the degree to which 

 the presence of traces of foreign metals influence the accuracy 

 of the results. 



My object in these notes is not to insist on any particular 

 application, but to bear testimony from a metallurgical point 

 of view to the delicacy and simplicity of the instrument 

 which Prof. Hughes has placed at our disposal ; and I would 

 offer him my sincere thanks for the liberal aid he has so readily 

 given me. 



VII. On the Theory of Faults in Cables. 

 By Oliver Heaviside. 

 1. ri^HE only kind of fault to be here considered is either a 

 J- local defect in the insulation, or an artificial connex- 

 ion between the conductor of a cable and the earth. When 

 a fault occurs in a submarine cable, its most manifest effect 

 on the working is to increase the strength of current leaving 

 the sending end, because the resistance is reduced ; while at 

 the same time the strength of current arriving at the distant 

 * Ann. de Chimie et de Phys. t. xxxix. 1828, p. 274. 

 f Instruction sur VEssai des Matieres df' } Argent par hi Yoic Humidc. 

 Paris, 1830. 



