246 THE ATOMIC WEIGHTS. 



Cleve* gives three values, but without data concerning 

 weiohinos or methods. Doubtless the oxide was converted 

 into sulphate, and the calculations were made with SO 3 = 



80: 



166.00 



i66.2i 

 166.25 



Mean, 166.153 



With SO 3 = 79.874, this becomes 165.891, and if only O 

 = 16, 166.273. These figures are undoubtedly the nearest 

 yet reached to the true value. According to Thalen,t who 

 reasons from spectroscopic evidence, the erbium of Hoeglund 

 was largely ytterbium. 



TERBIUM, SAMARIUM, PHILIPPIUM, DECIPIUM, THULIUM, 

 HOLMIUM, AND SORET's EARTH X. 



Concerning these substances, real or alleged, the data are 

 exceedingly vague. For phillippium Delafontainet gives 

 an atomic weight approximating to 123 or 125, and in the 

 same memoir decipium is ]:)ut at 171. It seems probable that 

 philippium may be identical with Cleve's holmiura and the 

 metal of Soret's earth X, while decipium comes near Cleve's 

 thulium, for which the discoverer gives a value of about 

 170.7.11 If decipium and thulium are identical, or if either 

 proves to be erbium or ytterbium contaminated with the 

 other, then we shall have a triad of metals with atomic 

 weights ranging from Er = 166 to Yb = 173, strikingly 

 parallel with lanthanum, cerium, and didymium. If we 

 take the natural arrangement of the elements as tabulated 

 after Mendelejeff's plan, somewhat modified in Roscoe and 

 Schorlemmer's " Treatise on Chemistry,§ we find that such 

 a triad should exist, and, furthermore, that another similar 



*Compt. Rend., 91, 382. 



t Poggend. Beiblatter, 5, 122. 1881. 



X Arch, des Sci. Phys. et Nat., Mars, 1880. 



II Compt. Rend., 91, 329. 1880. 



^ Vol. 2, Part 2, p. 507. 



