July 6, 1900.] 



SCIENCE. 



25 



oxids of this group, especially those of the 

 first triplet. 



The same may be said even more em- 

 phatically of the sulfids. Those of iron, 

 cobalt and nickel are fairly well investi- 

 gated, but of the remainder comparatively 

 little is known except the somewhat ex- 

 haustive work of Schneider on the thioplat- 

 inates and thiopalladates. After a very 

 considerable amount of work upon the sul- 

 fids of ruthenium, I have come to distrust 

 nearly all that has been published and to 

 have nothing definite to add myself. The 

 precipitates with hydrogen sulfid from ru- 

 thenium solutions (RuClj) contain appa- 

 rently a considerable amount of free sulfur, 

 but oxidize very rapidly with formation of 

 sulfuric acid on drying, making their compo- 

 sition very difficult of determination. Fi-om 

 ruthenate solutions a sulfid is precipitated 

 which seems to have the formula RuS^, 

 but there is no assurance that a part of 

 the sulfur may not be free and not com- 

 bined. 



Of all the compounds of the metals of the 

 eighth group, by far the best investigated 

 are those with the halogens, and upon our 

 knowledge of these rests the greater part of 

 our chemical knowledge of the platinum 

 metals. Yet here again our knowledge is 

 wholly inadequate. If we except the work 

 done under Wdhler's direction by Oppler 

 and Birubaum on the bromids and iodids 

 of iridium, that by Topsoe on the bromids 

 and iodids of platinum, we may say that 

 very little is known of any halids of this 

 group except the chlorids. In some in- 

 stances, as with ruthenium, even the chlo- 

 rids are very unsatisfactorily known. Of 

 nickel we know only the bichlorid ISTiClj ; 

 of cobalt the only stable chlorid is the bi- 

 chlorid, CoCl,, but the trichlorid, C0CI3 

 seems capable of existence in solution ; of 

 iron, the ferric chloride, FeClj, is the 

 stable compound, into which the ferrous 

 chloride, FeCl^, is readily oxidized. Here 



again the intermediate position of cobalt 

 is apparent. There is a strong tendency 

 on the part of all these chlorids to form 

 double salts, of which we have examples 

 in K„FeCl„ 3H,0, K.FeCl^, and Eb^FeCl,. 

 These salts seem to be broken up in so- 

 lution and the chlorin can be precipita- 

 ted by silver nitrate. Turning to the halo- 

 gen compounds of the platinum metals, 

 we find double salts of a very differ- 

 ent character. The common types for 

 platinum and palladium, for example, are 

 K^PtCl, and K^PtCl,. This latter type 

 seems also to be known for all platinum 

 metals except rhodium. Osmium, iridium, 

 and rhodium present also the type KjOsCl^, 

 while ruthenium and rhodium also form 

 salts of the type KjEuCl^. The most im- 

 portant features of these salts is that they 

 are not decomposed when dissolved in 

 water, silver nitrate precipitating not 

 silver chlorid alone, but the double chlorid 

 of the metal and silver; that is, for ex- 

 ample, when KjPtClj, is dissolved in 

 water, it is electrolj'tically dissociated, 

 K being the positive ion, while the nega- 

 tive ion is the group PtClg. The plat- 

 inum metal then in these salts is a part of 

 the negative ion. Double salts of this class 

 are, of course, well known, but nowhere 

 are they developed to the same extent as 

 in the eighth group, indeed double salts of 

 several acids are found among no other 

 metals. The question may be fairly raised 

 among the platinum metals as to whether 

 there is any salt which is electrolytically 

 dissociated giving the platinum metal as 

 the positive ion. 



The chlorids of ruthenium furnish an in- 

 structive illustration of the difficulties which 

 may arise in getting complete knowledge of 

 chemical facts as to what would naturally 

 be considered simple substances. As we 

 have seen, Glaus discovered ruthenium in 

 1844. He obtained two chlorids or rather 

 double chloi-ids, the oneK,EuClc and the 



