84 Dr. W. N. Hartley on tWe 



centre of the plate. Readings with this in a field of 459 

 c.G.s. units gave exactly Hall's (corrected) value for gold — 

 obtained in his case in a field of 1600 units. 



The above experiments, taken in connexion with Kundt's 

 work (loc. cit.), show that the Hall coefficient for gold is a 

 constant, at least for all fields between 12 and 21,500 units 



(C.G.S.). 



During the whole of the above investigation I have been 

 greatly assisted by the encouragement and advice of Professor 

 J. J. Thomson, to whom I desire to express my gratitude. 



Cavendish Laboratory, Cambridge, 

 July 30, 1901. 



VI. An Investigation into the Composition of Brittle Platinum. 

 By W. N. Haktley, D.Sc, F.R.S.* 



THREE years ago a small packet containing ten pin-points 

 of brittle platinum wire which had been in use for 

 dental purposes, with a similar number of points in another 

 packet labelled " tough Platinum " were submitted to me 

 with a request that I should examine the brittle platinum for 

 impurities of a metallic nature which it was supposed might 

 be present. Full details of the chemical investigation have 

 not been published, and as many points of interest have 

 arisen in connexion with the unusual method of examination 

 employed, I venture to present an account of it. Only a very 

 small quantity of defective material could be placed at my 

 disposal, the largest fragment of wire being only l/25th of an 

 inch or 1 millimetre in length. Their average weight was 

 under 20 mlgrs., and the whole parcel of them did not weigh 

 more than 0*2274 grs. When examined with the microscope 

 they were seen to possess a crystalline fracture f. The pins 

 broke readily on bending, and though they did not break 

 under the hammer the flattened wires had ragged edges 

 which was evidence of very imperfect malleability. It was 

 desirable, if possible, to examine each fragment separately as 

 there might be a difference in composition in pins which were 

 supplied by different makers, or which had come from 

 different sources. The very small quantity of material 

 available would have rendered any attempt to separate the 

 constituent substances by chemical means entirely futile, even 

 if the whole quantity of the metal were employed in a single 



* Communicated by the Author. 



f Andrews has investigated the macrocrystalline structure of platinum 

 bv etching plane surfaces, but pure platinum has not a marked crystalline 

 fracture. (Proc. Roy. Soc. vol. lxix. p. 433, 1902.) 



