142 Prof. H. Rose on the Composition of Samarskite. 



copper and iron I have subsequently continued on bars of larger 

 dimensions*. 



The length was 1180 millims., the breadth and thickness 

 35 millims. The results obtained agreed with the above; for 

 the copper bar there was obtained- 





k 







c8 



At the temperature 30*5 . 



. 66-80 



„ 339 . 



. 65-34 



41-0 . 



. 65-77 



41-8 . 



. 66-76 



„ 44-0 . 



. 65-13 



Mean 38*2 



65-96 



while the previous investigation gave 



At a temperature 51°'3 . . 64*66. 



These two values, reduced to the same temperature, are almost 

 identical, if it be assumed that the coefficient of temperature for 

 k has the same value for heat as for electricity. 



XVIII. On the Composition of Samarskite. 

 By Professor H. RosEf. 



THE numerous analyses of this remarkable mineral which 

 have been made in my laboratory do not agree very well 

 in their results. "Whilst M. von Peretz found in three analyses 

 14*16, 16*70, and 16*77 per cent, of oxide of uranium, the 

 amount of this oxide, according to Chandler, is 17*87 and 20*56 

 per cent. The former found 9*15, 11*04, and 8*36 per cent, of 

 yttria; the latter only 5*10 and 4*72 per cent. 



This want of agreement proceeds from the defective methods 

 employed in the separation of several of the constituents. The 

 separation of the peroxides of uranium and iron from the yttria 

 was effected by means of carbonate of baryta — a mode of separa- 

 tion of which I subsequently ascertained that it gives no certain 

 results, as it is difficult to avoid throwing down yttria together 

 with the precipitated oxides. This is the reason why the 

 amount of yttria appeared so small in Chandler's analyses. 

 The separation may, however, be well effected by means of 

 oxalic acid. 



When minerals containing niobium and tantalum are decom- 



* The details of these investigations will be given in the Nov. Act. Soc. 

 Upsal. S. 3. vol. iv. 



t Translated from the Monatsber. der Akad. der Wiss. zu Berlin, Nov. 

 1862, p. 622. 



