144 REPORTS ON INVESTIGATIONS AND PROJECTS. 



fore, dissolved in hydrochloric acid and reprecipitated with oxalic acid. 

 After converting into oxide this portion amounted to 290 grams. It was 

 dissolved in hydrochloric acid, the solution evaporated to dryness, the residue 

 taken up with water, poured into a mixture of 2.3 kg. of potassium sulphate 

 and 15 liters of water and stirred for four days. The solution obtained by 

 this process was filtered, precipitated with ammonium hydroxide, the precipi- 

 tate thoroughly washed and redissolved in hydrochloric acid. This solu- 

 tion of yttrite chlorides was then combined with the one obtained by the 

 carbonate process. 



In order to find out the approximate composition of the material prepared 

 in this way, an atomic weight determination was carried out by the well- 

 known method of Kriiss, as described by C. R. Bohm. 0.5126 gram oxide 

 gave 0.9852 gram sulphate. The atomic weight calculated from these figures 

 is 106. i§ for a trivalent element, a perfectly satisfactory number for crude 

 yttria. 



It seemed desirable, on account of the small quantity of material and its 

 high value, to undertake the first experiments with this crude yttria instead 

 of first extracting the single constituents from it. Thus far four electrolyses 

 have been run with the anhydrous chloride mixed with barium chloride, but 

 in these only pulverulent metal was obtained. The double fluorides, which 

 were also tried, did not give satisfactory results. The experiments will be 

 continued. 



The glucinum contained in the first filtrate from the rare-earth oxalates 

 and contaminated mainly with iron was purified by the usual methods with 

 ammonium carbonate and ammonium sulphide, and will be used for an 

 investigation of metallic glucinum. 



In connection with the experiments in which it was tried to reduce 

 thorium by means of metallic calcium, the behavior of this latter metal 

 towards ceria, zirconia, and their respective haloid salts was studied. Only 

 in one instance, by using anhydrous cerium chloride, mixing it with the 

 calculated amount of calcium filings and igniting with Goldschmidt's igni- 

 tion mixture, small globules of metallic cerium were obtained, but the heat 

 produced was not sufficient to melt both of the products of the reaction. In 

 all other cases only pulverulent metal was obtained. 



ON THE ATOMIC WEIGHT OF CAROLINIUM. 



The investigation on the elementary nature of thorium, published in pre- 

 vious Year Books, was criticized by Meyer and Gumperz. These authors 

 used, for the preparation of the anhydrous chloride, a method similar to 

 that described by Charles Baskerville, but which differs from it in certain 

 important details. Whereas the latter author found different values for the 

 atomic weights of the three fractions, Meyer and Gumperz claim that there 



