1896.] 



Constituent of certain Minerals. 



329 



Temperature, 14*28 ; pressure, 760*0 mm. 



Weight of 162*843 c.c. of gas 02957 gram. 



Weight of 1 litre at 0° and 760 mm 0-1911 



Density compared with = 16 2139 ,, 



A third weighing was carried out with a third separately collected 

 and purified sample. The spectrum showed the absence of nitrogen 

 and hydrogen. 



Temperature, 11*51°; pressure, 771*5 mm. 



Weight of 162-843 c.c. of gas 0*03024 gram. 



Weight of 1 litre at 0° and 760 mm 0*1906 



Density compared with = 16 2*134 „ 



Previous determinations of density have been recorded (' Chem. 

 Soc. Trans.,' 1895, p. 695). Some of these were made with a smaller 

 bulb, and the same degree of accuracy therefore cannot be claimed 

 for them. Still, they are included in the following table, so that the 

 record may be complete. 



Densities. 



Bulb of 33*023 c.c. capacity. Density, O = 16, 



Broggerite; heat alone 2152 



Broggerite; fusion with HKS0 4 . 2187 

 Cleveite; fusion with HKS0 4 . . . . 2*205 



The possible error in weighing might amount to one part in 60 or 

 70, which makes an uncertainty of 031 or 0*032 in the above 

 figures. 



Bulb of 162-843 c.c. capacity. Means. 



Mixture of gas from broggerite and cleveite. . 2'218 



Same mixture, re-circulated 2*133 



The last sample is probably the purer. 



Broggerite gas ; fresh sample 2*181 2*181 



Samarskite ; heat alone 2*121 



Samarskite ; fusion with HKS0 4 2*122 



Samarskite ; gas again circulated 2*117 



Samarskite ; after further circulation 2*114 



Fergusonite; heat alone 2*147 ~] 



Fergusonite; another sample 2*139 > 2*140 



Fergusonite; another sample 2*134 J 



Assuming a possible error of 0*2 milligram in weighing the globe 

 (and this error is certainly too high), the error might amount to 

 0014. 



It appears to me that these numbers furnish some ground for the 



2*118 



