24 



SCIENCE. 



[N. S. Vol. XII. No, 



tive series of the elements of the sixth and 

 seventh groups; viz, CrO^ SlbO,, WO3, UO3, 

 MnO.,. This type is not represented in the 

 second or third triplet of group eight. Po- 

 tassium ferrate, K^FeO^, exists only in solu- 

 tion and is very unstable; potassium ru- 

 thenate, K^EuO^ is stable when dry but 

 slowly decomposes when in solution; potas- 

 sium osmate, K^OsOj, on the other hand, 

 has a very considerable degree of stability. 

 Of the lower base-forming oxids, iron has 

 not only the sesquioxid, Fe^Oj, and mon- 

 oxid, FeO, but also several intermediate 

 oxids which may be looked upon as merely 

 compounds of these two — such is magnetite. 

 In the case of ruthenium, the sesquioxid, 

 EUjOj, seems to be what one might call the 

 normal base-forming oxid. The different 

 conditions which occasion the formation of 

 the lower oxids of osmium are ill known, 

 though several different oxids seem to exist, 

 as OsO, OSjOj, and OsOj. Much more in- 

 terest, however, attaches to the tetroxids of 

 ruthenium and of osmium, RuO^ and OsO^, 

 which are the highest volatile oxids of any 

 known element. The almost intolerable 

 odor of osmium tetroxid incited Tennant, 

 in 1803, to give its name to this element, 

 while ruthenium tetroxid, first noticed by 

 Claus, has, if not too concentrated, a rather 

 fresh pleasant odor, with just a suspicion of 

 the smell of ozone, due probably to the for- 

 mation of ozone in the decomposition of the 

 oxid. As far as physical properties go, 

 these oxids, while solid at ordinary tem- 

 perature, melt easily and can be distilled. 

 Euthenium tetroxid is, however, far less 

 stable than the corresponding osmium oxid, 

 for it decomposes slowly at ordinary tem- 

 perature, and explodes with great violence 

 if heated much above 105°. On one occa- 

 sion Deville and Debray* attempted to 

 distil a hundred and fifty grams of ruthen- 

 ium tetroxid, and when the temperature 

 reached a little above 100°, the whole mass 

 *Ann. Chim. Pliys. [5], 4, 537 (1875). 



exploded with terrific violence, filling the 

 laboratory with a dense sooty smoke. In a 

 recent similar explosion in my own labora- 

 tory, occasioned by the contact of a little 

 alcohol with ruthenium tetroxid, but hap- 

 pily on a much smaller scale than that of 

 Deville and Debray, this black soot was 

 found to be readily soluble in hydrochloric 

 acid. This is unexpected, as all the an- 

 hydrous lower oxids of ruthenium are in- 

 soluble in acids, and from its methods of 

 formation the black substance could hardly 

 be anything other than an anhydrous oxid 

 or the metal itself. Osmium tetroxid is 

 commonly known as osmic acid, but as a 

 fact these tetroxids are neither acid-form- 

 ing oxids nor are they peroxids in the or- 

 dinary acceptance of the term. When 

 treated with an alkali, we have a gradual 

 reduction with formation of perruthenate 

 and ruthenate, or of osmate. 



Turning to the third triplet, we have the 

 monoxid of nickel well characterized, and it 

 is, we may say, the only well-characterized 

 oxid of the metal, for though higher hy- 

 drated oxids of nickel exist, and perhaps 

 anhydrous oxids also, their composition is 

 not definitely known. With palladium and 

 platinum also, monoxids seem to exist and 

 dioxids as well (PdO, and PtOJ. Plati- 

 num dioxid may be looked upon as being 

 perhaps a very weak acid- forming acid. 

 While nickel forms practically only the 

 monoxid NiO, and ii'on forms from choice, 

 as we might say, the sesquioxid, Fe^O.,, the 

 intermediate metal cobalt, while forming 

 most generally the monoxid, is easily oxi- 

 dized to the sesquioxid, Co^Oj, and thus 

 cobalt may be considered in its relations to 

 oxj'gen, as intermediate between nickel and 

 iron. Similarly in the case of rhodium and 

 iridium, there is a strong tendency to form 

 the sesquioxid, so that this middle triplet 

 is intermediate between the other two trip- 

 lets of the group. As a whole, there is a 

 large field at hand in a revision of the 



1 



