OF ARTS AND SCIENCES. 265 



II. (25) 0.3563 gr. gave 0.1114 gr. C 2 8 = 31.26 per cent. 



(26) 0.3500 gr. « 0.1091 gr. " =31.20 



(27) 0.6848 gr. « 0.3317 gr. R 2 3 = 48. 1 :i 



(28) 0.6664 gr. " 0.3223 gr. " =48.37 



Atomic mass 143. '■•■'>. 



III. (29) 0.3500 gr. gave 0.1113 gr. C 2 8 = 31.78 



(30) 0.3127 gr. " 0.09936 gr. " =31.77 « 



(31) 0.6859 gr. " 0.3345 gr. R 2 8 = 48.75 » 



(32) 0.7087 gr. " 0.3447 gr. '• =48.65 



Atomic mass 141.55. 



IV. (33) 0.3350 gr. gave 0.1042 gr. C 2 8 = 31.12 



(34) 0.3410 gr. " 0.1060 gr. " =31.09 " 



(35) 0.6369 gr. " 0.2958 gr. 1L0 3 = 46. 1 1 



(36) 0.6468 gr. « 0.3004 gr. " =46.45 « 



Atomic mass 137.25. 



V. (37) 0.2042 gr. gave 0.06864 gr. C 2 3 = 33.62 « 



(38) 0.2031 gr. " 0.06828 gr. " = 33.62 " 



(39) 0.3045 gr. " 0.1409 gr. R 2 3 = 46.26 



(40) 0.3400 gr. "0.1569 gr. " =46.14 " 



Atomic mass 124.40. 



The filtrate from the last portion of oxalates contained only traces 

 of earths. The analyses show that the earths with the highest atomic 

 masses are precipitated first by oxalic acid. The average rate of de- 

 crease of mean atomic mass is five units for each operation, but the 

 rate is by no means uniform. The results prove that precipitated oxa- 

 lates must be very carefully mixed mechanically before analysis when 

 more than one earth is present. The analyses also show that four 

 fifths of the earths might have been precipitated at once by oxalic 

 acid if the object had been to obtain yttria with the least outlay of 

 time and labor, the atomic mass of the last fifth being 124.40. 



Oxychlorides. — When the chlorides of the metals belonging to the 

 cerium and yttrium groups, or, more generally stated, of the m< 

 yielding the rarer earths, are carefully heated, oxychlorides are formed 

 in greater or less proportion mixed with undecomposed chlorides. We 

 may distinguish these as acid and basic chlorides. Water dissolves 

 out the former very readily, and leaves nearly or quite colorless basic 

 chlorides, which are, relatively at least, insoluble. I find that in this 

 way the earths present an- separated into two groups, and thai by 

 repeating the operation upon each group a second differentiation ifl 



