May 15, 1903.] 



SCIENCE. 



775 



which were subsequently appropriated by 

 samarium, reported as a constituent of 

 didymium by de Boisbaudran. Samarium 

 would now have the name of decipium, but 

 for the fact that, in 1881, Delafontaine 

 declared his decipia could be resolved into 

 an oxide without absorption spectrum (true 

 decipia) and one with these lines, or 

 samarium. 



J. Lawrence Smith, of Kentucky, in the 

 seventies, announced mosandrum in samars- 

 kite. Marignae and Delafontaine inde- 

 pendently pointed out that mosandrum was 

 the same as terbium, while later de Bois- 

 baudran demonstrated that it was a mix- 

 ture of terbia and gadolinia. This 'nebula 

 of elementary matter,' as Petterson puts 

 it in that charming account of the life work 

 of Nilson, appeared to clear up through 

 the work of the English, French and Swiss 

 chemists, Roseoe, de Boisbaudran and 

 Marignae. While 'the beginning of crea- 

 tion is light,' as Carlyle says, the millen- 

 nium has not yet arrived, for the earths 

 obtained from gadolinite began ■ to break 

 up into a number of new earths. 



Cleve (1873) found that the bands of 

 erbiiun with an atomic weight of 170.5 

 could be split into those belonging to one 

 element forming a red oxide with the char- 

 acteristic emission spectrum (by incandes- 

 cence) of old erbium and another group 

 of two absorption bands in the visible spec- 

 trum. These were shown to belong to 

 thulitun. 



Five years later Marignae found all the 

 absorption bands could be eliminated by 

 successive fractioning, whilst the atomic 

 weight of the remaining oxide increased. 

 This oxide gave colorless salts without ab- 

 sorption bands, and the name jrtterbium 

 was assigned to it, with an atomic weight 

 of 172.5. In the erbia fractions Soret 

 foimd bands which could not be attributed 



to erbium. This body, designated X, sub- 

 sequently proved to be Cleve 's holmium. 



Material giving out, Marignae, with the 

 true scientific spirit, begged other and 

 younger men to take up the work, using 

 larger amounts. This Nilson did and veri- 

 fied Marignae 's work. Just before reach- 

 ing the same point Marignae arrived at, 

 however, Nilson obtained a nitrate of a less 

 basic material of lower atomic weight. One 

 fraction continued to drop, while the other 

 rose imtil, in the year following (1879), 

 assisted by Thalen, who examined the 

 products with the spectroscope, Nilson 

 separated probably the two best defined of 

 the rare earths, scandium (44.1) and ytter- 

 bium (173). Nilson showed the location 

 of these elements in the Mendeleeff table, 

 the properties of the former having been 

 predicted. 



Referring to these elements Mendeleeff 

 says: 'These metals which are rare in na- 

 ture, resemble each other in many respects, 

 always accompany each other, are with dif- 

 ficulty isolated from each other and stand 

 together in the periodic system of the ele- 

 ments. ' The last statement is based largely 

 upon analogy, a most valuable method of 

 argument in scientific generalizations with- 

 out doubt, but, as Davy once said : ' Analogy 

 is the fruitful parent of error.' 



In 1880 Marignae attacked samarskite, 

 and by fractioning the double potassium 

 sulphate obtained two oxides in almost pure 

 state, as follows : 



Ya giving a white oxide, colorless salts 

 and no absorption bands. Six years later 

 it was called gadolinium and the atomic 

 weight 156 assigned it by Marignae, de 

 Boisbaudran, Cleve and Bettendorff. 



Y/S proved to be samarium of de Bois- 

 baudran, or Delafontaine 's original decip- 

 iiun. Marignae, Cleve, Brauner and Bet- 

 tendorff determined its atomic weight ( 149— 

 150). While the elementary character of 



