142 REPORTS ON INVESTIGATIONS AND PROJECTS. 



fluoride with aluminum under suitable conditions, and subsequent treat- 

 ment of the regulus obtained with caustic alkali to remove the excess of 

 aluminum. The metal prepared in this way, however, was a crystalHne 

 powder which still contained some impurities. The removal of these im- 

 purities was absolutely essential for the subsequent determinations of the 

 physical and chemical properties of the metal. The only effective way for 

 accomplishing this is remelting it repeatedly. But this method can not be 

 applied to thorium under ordinary conditions, because the melting point of 

 this metal is extremely high and because it is attacked by all inert gases, 

 forming hydride, nitride, or carbide. It can not be fused under a cover of 

 an alkaline chloride, because the metal oxidizes at a temperature much lower 

 than the melting points of the chlorides. It is necessary, therefore, to fuse 

 it in an absolute vacuum. We hope to carry out these experiments in a 

 vacuum resistance furnace of the Arsem type. 



In the meantime some other methods were tried with the view of obtaining 

 the fused metal directly. Muthmann and his pupils showed that the metals 

 of the cerite group may be obtained by electrolysis of the fused anhydrous 

 chlorides in a water- jacketed copper crucible, using a very thin carbon rod as 

 cathode and a thick carbon anode. In order to find out the most favorable 

 conditions, samples of metallic cerium, lanthanum, and neodymium were 

 prepared with satisfactory results. It was further tried to obtain these 

 metals from their respective double fluorides with potassium by electrolysis. 

 The double fluorides were prepared in the following way : Concentrated 

 solutions of the chlorides or nitrates were heated on the water-bath and pure 

 hydrofluoric acid added. The gelatinous precipitate obtained at first soon 

 changed into a fine, heavy, sandy powder which could be easily filtered by 

 means of a hard rubber funnel and washed with water. It was then dried 

 at 1 20° C. and finally fused with 3 molecules of potassium fluoride. When 

 working under the same conditions as Muthmann, these double fluorides 

 were readily fused by the current used, but in no case was metal formed, as 

 Muthmann had also observed in the case of didymium. By changing the 

 proportions of rare-earth fluoride to potassium fluoride the melting point and 

 the resistance of the substances was also altered. By electrolysis of these 

 bodies metal could not be prepared. 



After these preliminary experiments the preparation of metallic thorium 

 was attempted. As the fluoride of this element is prepared much more 

 readily than the anhydrous chloride, the investigation was begun with the 

 double fluoride of thorium and potassium made in exactly the same way as 

 those of the cerite metals, mixing two molecules of potassium fluoride with 

 one of thorium fluoride. The material fused very readily in the electrolytic 

 cell ; however, no metal was obtained. Next to this experiment the elec- 

 trolysis of double chlorides of thorium was attempted. Anhydrous thorium 

 chloride was prepared in a porcelain tube by Matignon's method, i. e., by 



