July 6, 1900.] 



SCIENCE. 



21 



out in a single group ; dumped into a chem- 

 ical Gehenna as it were, while the rest of 

 the elements were reduced to orderly ar- 

 rangement. Lothar Meyer, however, saw 

 that there was a possibility that these ele- 

 ments also might be amenable to system, 

 and under his direction Carl Seubert began 

 the revision of the atomic weight of irid- 

 ium.* This he found to be more than four 

 units less than the figure formerly used, 

 and now the order of these elements ap- 

 peared to be iridium, gold, platinum, os- 

 mium. Three years later Seubert f revised 

 the atomic weight of platinum, finding it 

 lower than that of gold, and this work Was 

 confirmed by Halberstadt, J and by Ditt- 

 mar and McArthur.§ The only anomaly 

 in these four metals now was in osmium 

 and this also was resolved by Seubert, 1 1 who 

 found that the old value of Berzelius and 

 Fremy was about eight units too high, and 

 that so far from having an atomic weight 

 greater than that of gold, osmium in reality 

 has the lowest atomic weight of the four 

 metals. This revision was justly accounted 

 a great triumph for the periodic law. 



As with the other metals, so also much 

 doubt existed as to the atomic weights of 

 rhodium and ruthenium, but the work of 

 Seubert and of Joly, while changing some- 

 what the older figures, confirmed the order 

 given in Meyer's table. Much work has 

 been done on palladium by Keiser and by 

 Keller and Smith in this country, by Bailey 

 and Lamb and by Joly and Leidie abroad. 

 The figures for platinum and palladium rep- 

 resent a much greater degree of accuracy 

 than those for the other four platinum met- 

 als. Indeed it must be said that little accu- 

 racy can be claimed for the present figures 

 of rhodium and iridium, and certainly those 



*Inaug. I)is3., Tubingen, 1879. 

 t Ber. d. Chem. Ges., 14, 865 (1881). 

 tlhid., 17 (1884), 2962. 

 • iJ. Soc. Chem. Ind., 6, 799 (1887). 

 II Ber. de Chem. Ges., 21, 1839 (1888). 



of ruthenium and osmium cannot be de- 

 pended on more closely than half a unit. 



In the case of the three other metals of 

 this group, the atomic weight of iron has 

 been well determined, but is now being 

 subjected to a most careful examination in 

 the laboratory of Professor Theodore Rich- 

 ards. As was supposed a few years ago to 

 be the case with iridium and platinum, so 

 cobalt and nickel were thought to have the 

 same atomic weight. Then Lothar Meyer 

 showed that, judged by their properties, 

 nickel should follow cobalt in the periodic 

 system and hence have the higher atomic 

 weight. Eevisions of these metals followed, 

 but the more accurate the work the more 

 probable it appeared that the atomic weight 

 of nickel is below that of cobalt. It was 

 suggested that nickel was quite probably a 

 mixture and efforts were made to resolve it 

 into its constituents. In this connection 

 will be recalled the eiforts of Gerhardt 

 Kruss to decompose nickel, in which for 

 some time he thought he had been success- 

 ful and christened the new metal gnomium. 

 But like so many other aspirants for chem- 

 ists' favor, gnomium proved to be but a 

 mixture. The latest work on these metals 

 by Richards and Cushman and Baxter, far 

 surpassing all that has previously been 

 done, confirms the higher atomic weight of 

 cobalt, and lends no support to the view 

 that nickel is anything but a simple ele- 

 ment. 



Here we meet apparently one of those 

 chemical mysteries, which seem to baffle 

 our attempts at solution. We are not per- 

 mitted to doubt the correctness of the gen- 

 eral principles of the periodic system, and 

 yet here, and the case is perhaps not unique, 

 two elements seem to have exchanged places. 

 "When we know lohy the properties of an 

 element are a function of its atomic weight, 

 we shall perhaps come to understand why 

 the atomic weight of nickel is not greater 

 than that of cobalt. 



