| 
: 
; 
PRESIDENTIAL ADDRESS. 415 
not; Kekulé’s hypothesis requires addition to precede substitution. Is there any 
experimental evidence to indicate where the balance of probability lies? I think 
it can be argued that the phenomena of substitution observed with optically 
active substances do not lend support to the views of Arrhenius or of Nef, which 
imply actual or virtual dissociation, but that they point to the intermediate 
formation of an additive product, which undergoes scission as Kekulé supposed. 
Such an additive product can be formed only if residual valencies be present in 
both carbon compound and addendum. 
The argument runs thus: Unless valency has fixed direction in space, a con- 
ception now abandoned if modern theories of valency be accepted, the conclusion 
seems to be inevitable that dissociation of the optically active compound :— 
CWXYZ into CWXY + Z, 
must lead to racemisation, the radicals W, X, Y, distributing themselves in two- 
dimensional space, thus destroying the asymmetry; whence it follows that 
introduction of the substituent, V, into the molecule in place of Z can give 
rise only to an optically inactive product. Now, it is a well established fact 
that a radical attached directly to the asymmetric carbon atom may be replaced 
by another without racemisation following.*® Therefore, preliminary dissocia- 
tion being excluded, Kekulé’s additive hypothesis remains. But the prolonged 
study of that remarkable reaction known as the ‘ Walden inversion’ by Emil 
Fischer, McKenzie, and other investigators has led to results which, if the views 
formed independently by Fischer,*® Werner,*’ and Pfeiffer ** may be accepted, 
are inexplicable unless a preliminary addition, effected as it is supposed by means 
of residual valencies, precedes this replacement of the eliminated radical by the 
substituent. 
The Walden inversion may be illustrated by a brief statement of some of the 
facts discovered in connection with the conversion of optically active chloro- 
succinic acid into malic acid 
R + CH(OH) *CO,H — R- CHCl: CO.H — R:CH(OH):CO,H. 
Walden found that (evo-chlorosuccinic acid, obtained from dextro-malic acid, 
furnished either dextro- or Jevo-malic acid, according to the reagent used to 
effect of the replacement of the Cl by the OH radical. 
se -,-7( + Ag,O )—> I-malic acid 
1-chlorosuccinic acid Sait a i ae eee 
And as the corresponding inversion was found to occur with dextro-chlorosuccinic 
acid under similar conditions, a complete cycle of changes can be brought 
about.*° That preservation of optical activity, and not racemisation, should 
accompany the replacement of a radical, attached to the asymmetric carbon atom, 
by another is a fact of much theoretical interest, as has already been indicated ; 
that a change in the sign of rotation should occur when an exchange of the same 
radicals is achieved by one reagent and not by another is a mystery, that deepens 
rather than diminishes with each addition to the list of inversions, already long, 
in which it has been observed.*® In all probability the discovery of the Walden 
29 P. Walden, Ber., 1895, 28, 1297; W. A. Tilden and B. M. C. Marshall, 
Trans. Chem. Soc., 1895, 67, 494. 
26 H. Fischer, Annalen, 1911, 381, 123. 
*7 A. Werner, Ber., 1911, 44, 873. 
*8 P. Pfeiffer, Annalen, 1911, 883, 123. 
2» P. Walden, Ber., 1896, 29, 133; 1897, 80, 3146; 1899, 32, 1833, 1855. 
*° Without the aid of a model it is not possible to show that the production 
of the'dextro- or Icevo- acid may be accounted for by the hypothesis that an 
intermediate additive compound is formed, which undergoes scission in one or 
other of two ways. Diagrams of models will be found in Fischer’s paper 
(loc. cit. cf. Annual Reports on the Progress of Chemistry (Gurney and Jackson) 
> 
