504: SMITITSOXIAX MISCELLANEOUS COLLECTIONS VOL. ')A: 



This mean agrees very closely with the figures given by Jones. The 

 early determinations, previous to Marignao. might be properly rejected 

 altogether, as their influence upon the combination is imperceptible. 



For the quantity of Lag (804)3 proportional to 100 parts of BaSO+, we 

 have five experiments, which may be given equal weight and averaged 

 together : 



Marignac 81.022 



Mari{?nac 80.934 



Rammelsberg 79.276 



Czudnowicz 79.117 



Erk 79.921 



Mean, 80.054, ± .270 



In all there are eight ratios from which to calculate the atomic weight 

 of lanthanum : 



(1). Percentage of LaA in La„(S0,)3, 57.5469, ± .00034 



(2). 3BaCL:La,(SOj3: :100:91.322, ± .048— Marignac 



(3). 3BaS04:La.(SOj3:: 100: 80.054, It .270 



(4). 3BaS04:LaA: : 100: 46.671, ± .075— Holzmann 



(5). Percentage of LaoO., in iodate, 23.447, ± .0216 — Holzmann 



(6). Percentage of La^O^ in magnesian nitrate, 21.3056, ±: .058 — Holzmann 



(7). 3aO.,:La,03: : 100: 151.246, ± .0400 



(8). 0:La: : 16: 139.09, ± .0430— Feit and Przibylla 



Hermann's single experiment on the carbonate is omitted from this 

 scheme as being of no value. 



The antecedent atomic weights are- — 



Ba = 137.363, ± .0025 CI = 35.4584, ± .0002 



Mg = 24.304, ± .0006 I = 126.9204, ± .00033 



S = 32.0667, ± .00075 N = 14.0101, ± .0001 



C = 12.0038, ± .0002 H = 1.00779, ± .00001 



Hence, 



From ratio 3 La = 136.205, ih .6875 



"5 137.652, ±.1557 



" 6 138.637, ± .4587 



" 1 138.801, ± .0020 



" 8 139.090, ±: .0430 



" 7 139.363, ± .0303 



" 4 139.416, ± .2626 



" 2 141.208. + .1500 



It is evidently unnecessary to combine these values into a general 

 mean, for only one of them, that from ratio 1, carries any appreciable 

 weight. The other values could not modify it to any noteworthy extent. 



