86 v Dr. W. N. Hartley on the 



would afford satisfactory proof of its presence when so small 

 a quantity of material is available for analysis. Carbon also 

 renders platinum brittle. The platinum cones or caps used 

 in the ignition of the mixture of gas and air, or petroliiie 

 vapour, which is used in gas and motor-car engines, become 

 so brittle that they may be broken up between the fingers. 



The final conclusion was that some non-metallic impurity, 

 either carbon or phosphorus, is the cause of the, brittle and 

 markedly crystalline character of these pins, as observed in 

 their fractures. It appeared possible that either of the sub- 

 stances might be introduced in the working up of old platinum. 

 Both phosphorus and carbon can be eliminated by prolonged 

 fusion of platinum under the oxy- hydrogen blowpipe flame ; 

 accordingly a number of the brittle pins were fused, and 

 under this treatment the metal became greatly improved in 

 malleability, which fact affords a strong confirmation of the 

 correctness of the conclusion arrived at. I have been informed 

 that when this matter was communicated to the International 

 Congress at Paris in the year 1900, by the Dublin Branch of 

 the Dental Association, it elicited a statement to the effect 

 that the electric furnace has latterly been used for fusing 

 platinum. This method of fusion would unquestionably lead 

 to the introduction of phosphorus, and probably also of carbon 

 and silicon, from the metal being in contact with carbon 

 electrodes. We know that phosphorus is introduced into 

 calcium carbide by the ash of the powdered coke used ; and 

 from the carbon electrodes in which silicon is always present, 

 this element may possibly enter into platinum. 



Experimental Details. 



There were difficulties encountered in the examination of 

 the spark-spectra emitted by these small fragments of metal. 

 In the first place they could not be submitted to fusion lest 

 they might be altered in composition, and secondly, it was 

 very desirable that they should be examined separately. The 

 largest fragment as already mentioned was only 1 mm. long, 

 and some of them were no more than half this length. 

 Conceive the difficulty of getting a pure spectrum from 

 this. If held in a screw clip, the metal of the clip would 

 contribute extraneous matter to the spark, steel for instance 

 w 7 ould furnish iron, while graphite or silver would at once 

 constitute itself an electrode in place of the platinum because 

 either of them is a much superior conductor. 



The usual spring or screw clips were in fact found to be 

 useless. Mounting was accomplished in two ways, first by 

 fixing the pin in a clip of platinum-wire which, however, was 



