28 



SCIENCE. 



[N. S. Vol. XII. No. 



ditions, whicli happily Professor Antony- 

 lias hit upon. 



But while the platinum metals seem to 

 form few simple salts, few or none show 

 such a decided tendency to form double 

 and complex salts, and this property is, to 

 some extent, shared by the three light 

 metals of the group. 



Best known and best developed of these 

 compounds are the cyanids, which are es- 

 pecially familiar to us in the prussiates of 

 iron. In nickel we have the ordinary 

 cyanid, K,M(CN), or 2KCN, MCCE'),, 

 formed by the solution of nickel cyanid in 

 potassium cyanid. As electrolytically dis- 

 sociated, the nickel is a positive ion, and 

 the double salt is at once broken up by 

 acids with the precipitation of nickel cyanid. 

 The double cyanid of palladium, K^Pd 

 (ClSr)^ is similar but less easily decomposed. 

 The corresponding double cyanid of plati- 

 num, K,Pt(CiSr)^ is clearly a salt of the 

 complex platinocyanic (or cyanoplatinous) 

 acid, H,Pt(CN')j, which is formed on treat- 

 ing the salt with a strong acid, can be 

 separated in a pure condition, and is an 

 acid strong enough to expel hydrochloric 

 acid from sal ammoniac. The platinum 

 atom is here a constituent of the negative 

 ion, Pt(CN),. 



If we proceed from nickel along the hori- 

 zontal series, we find that while a double 

 cobalt cyanid, K,Co(CN")„ or 4 KCN",- 

 Co(CN)j, can be formed, it is verj' unstable, 

 and belongs to the same easily decompos- 

 able class as the double nickel cj^anids. 

 This cobalt cyanid has, however, a great 

 tendency to oxidize and form potassium co- 

 balticyanid, K3Co(CN")5, which is stable 

 and a salt of the cobalticyanic acid, which 

 can be obtained in a free state. In passing 

 we note a very interesting point, that under 

 the influence of such reducing agents as 

 potassium cyanid, potassium nitrite, and 

 potassium sulphite, cobalt shows a great 

 tendency to become oxidized from its biva- 



lent condition to the very stable complex 

 compounds in which it is trivalent ; under 

 other circumstances, simple compounds in 

 which cobalt is trivalent are formed with 

 great difficulty and are of decided instabil- 

 ity. This seeming anomalous property still 

 demands an explanation. 



Turning to the iron cyanids we find both 

 types, K,Fe(CN), and KjFeCCN)^, ferro- 

 cyanid and ferricyanid, well developed 

 and extremely stable. From each, the cor- 

 responding acid can be obtained in a free 

 state, and is a strong acid. Of the remain- 

 ing metals, the dou^ble cj'anids of rhodium 

 and iridium resemble the cobalticyanid, 

 while of iridium the iridocyanid, K^Ir (CN)g 

 is also known, and is stable, thus complet- 

 ing the analogj' found in the nickel group. 

 Potassium ruthenocyanid, E^Eu (ON), and 

 osmocyanid, K^Os(CN)j resemble the ferro- 

 cyanid, the free acids being easily separable 

 from the salts. Outside of the eighth group, 

 the stable complex cyanids are known only 

 in the case of maugauese and chromium. 



Eegarding the constitution of the double 

 C3'anids, you are all familiar with the vari- 

 ous suggestions that have been made from 

 time to time, which involve the polymeriza- 

 tion, probably by threes, of the cyanogen 

 group. To this there have been raised two 

 objections : an explanation which is satis- 

 factory for the double cyanids should also 

 be available for the double chlorids, as 

 KjPtClj which are also salts of complex 

 acids, and where polymerization by threes 

 is at least improbable ; and second it is pos- 

 sible to replace a single cyanogen group or 

 chlorin atom, without changing essentially 

 the nature of the molecule, as in sodium 

 nitroprussid, ]Sra2Fe(C]Sr)5]SrO, and potas- 

 sium nitrosochlorruthenate, K^EuCljlSrO. 

 There is a large field for study in these cya- 

 nids from the standpoint of the newer 

 physical chemistry. 



Closely connected with the chemistry of 

 the cyanids is that of the thiocyanates, but 



