NA TURE 



433 



THURSDAY, MARCH 8, 1906. 



A REVISED DOCTRINE OF VALENCY. 



Neitere Anschauungen auf dem Gebiete der anor- 

 ganischen Chemie. By Prof. A. Werner. Pp. xii + 

 1S9. (Brunswick : F. Vieweg and Sohn, igoj.) 

 Price 5 marks. 

 T N 1893 Prof. Werner published his first important 

 paper on the constitution of inorganic compounds, 

 and expounded a new theory for the classification of 

 the large and complex group of substances known as 

 metal-ammonia compounds. Of these compounds a 

 few are common enough, such as the deep blue sub- 

 stance formed by the addition of ammonia to solutions 

 of copper salts, but the greater number do not come 

 within the range of ordinary analytical chemistry; 

 they are for the most part of no industrial importance, 

 and consequently they are scarcely heeded except by 

 a very limited number of chemical workers. The 

 class of cobaltamines which has vexed many a gener- 

 ation of chemical students does not measure the limits 

 of complexity to which these compounds extend, and 

 certainly, without the guiding light of some good 

 theory as to their structure, the metal-ammonia group 

 constitutes one of the most bewildering tracts of 

 inorganic chemistry. 



It is therefore a real service that Prof. Werner 

 renders in publishing in this volume (No. 8 of a series 

 of monographs on natural and mathematical science, 

 collectively called " Die Wissenschaft ") a full ex- 

 position of his theory and an apercu of the compounds 

 to which it specially applies. We must not give the 

 impression, however, that Werner's theory applies 

 only to metal-ammonia compounds. It is a general 

 theory ranging over chemistry as a whole, and is 

 applied to so simple a substance as sulphuric acid. 

 To describe it in a few words is almost impossible. 

 Perhaps the most central thing is the substitution 

 of the idea of association (Anlagerung) for the strict 

 and definite linkage implied in the ordinary valency 

 theory. This idea has, in a vague way, long prevailed 

 in chemistry in the distinction drawn between atomic 

 and molecular compounds. In CuS0 4 we are accus- 

 tomed to represent definite atomic linkages according 

 to the definite valencies of the component atoms. How 

 are we to represent the attachment of 5H 2 to CuS0 4 

 in the hydrated salt? The idea that valency is not a 

 sharply fixed quantity, and that it is not necessarily 

 exhausted when a stable compound is formed, is 

 familiar to us in the hypothesis of residual affinity. 



Prof. Werner would have us revise our somewhat 

 diagrammatic and artificial ideas of valency. We are 

 to think of an atom as a small material sphere, from 

 the centre of which the attractive force of affinity is 

 exerted uniformly in every direction. Segregation of 

 the affinity into units of valency is not to be supposed; 

 valency is merely to mean the observed proportional 

 numbers in which atoms associate with one another. 

 It is not dependent on one atom alone, but on the 

 nature of all the atoms that form the molecule. The 

 NO. 1897, VOL - 73] 



proportion of the affinity which is spent between two 

 atoms is confined to the restricted circular surface of 

 contact [theoretically one would say a point], the 

 Bindeflache, and depends in a high degree on the 

 nature of the atoms. Thus, the author adds, we 

 obtain a rational picture to represent the varying 

 valency of an atom, whilst from the dependence of the 

 distribution of the surfaces of contact on the relative 

 magnitudes of the atoms we reach, without further 

 hypothesis, a space configuration of the molecule. 



As the simplest possible illustration of the advan- 

 tage claimed for these views we may cite the 

 formation of sulphuric acid by the union of S0 3 and 

 H 2 0. In each of these molecules it is usually con- 

 sidered that the valencies of the atoms are satisfied. 

 That being so, the readiness of the two molecules to 

 unite must be attributed to some selective action. 

 This is found in the tendency for the group (OH) to 

 be formed, and hence we write S0 3 + H 2 = 2 S(OH) 2 . 

 According to Prof. Werner, on the other hand, we 

 must suppose that neither the sulphur in S0 3 nor the 

 oxygen in H,0 has spent its affinity, and that 

 accordingly the combination of the two molecules is 

 to be represented as follows: 3 S + OH 2 = 3 S.OH 2 . 

 At the same time he admits that the compound may 

 pass into the configuration 2 S(OH) 2 . It might seem, 

 then, that there is not much gained. But he claims 

 that the great merit of this view lies in its conformity 

 with that which must be taken of analogous com- 

 binations where a secondary arrangement does not 

 take place. Such a case is to be found in the union 

 of halides to form what are, though commonly called 

 double salts, compounds strictly of the same order 

 as oxysalts. Thus we have KCl + AuCl 3 = KAuCl. l 

 analogous to K,0 + S0 3 = K 2 SO. l . KAuCI 4 is as 

 much a potassium salt as K 2 S0 4 , and, though it may 

 not be impossible to give an ordinary valency formula 

 to some of these compounds, their formation cannot 

 be explained by anything corresponding to the sup- 

 posed primary cause (the formation of KO groups) 

 in the union of K 2 and S0 3 . The attempt to bring 

 double halides as a whole within the ordinary valency 

 doctrine has not been successful. 



But, as stated above, Prof. Werner's theory has 

 arisen in connection with the metal-ammonia com- 

 pounds, and we will conclude this notice with a slight 

 indication of its application there. In the metal- 

 ammonia compounds we have an electropositive atom, 

 a number of (NH 3 ) groups, and electronegative atoms 

 or groups. It is supposed that the positive atom has 

 the power of associating itself with or coordinating a 

 certain number of atoms or groups which must be 

 supposed to be in contact with this central atom and 

 to constitute a sort of first layer. The whole group 

 also forms the positive ion. Beyond and outside this 

 we have negative atoms or groups which give the 

 negative ions. 



Luteo-cobalt chloride has the composition 



CoCl 3 6NH 3 . 



Cobalt is here shown with the valency of a triad, and 



as a matter of fact the whole of the halogen may be 



precipitated by silver nitrate. Purpureo-cobalt chloride 



U 



