81 



First of all the configuration may be deduced fVoiu properties 

 that are to be anticipated in consequence of the reciprocal inthience 

 of the groups in the molecule; this is tiie surest way, if enough 

 (critically examined) data for comparison be available. 



Among these properties are: dissociation constants of acids, for- 

 mation of anhydride, resolution into optical antipodes, etc. 



The formation of complex compounds is to be included in so far as the valencies, 

 which are the bearers of cistrans isomerism, are not attacked. 



This is seldom easily established, since the structure of complex compounds is 



uncertain. Probably we are authorized to use this method in judging the cis-trans 



isomerism of cyclic glycols by the influence on the conduclivity of boric acid, 



I because the boric acid-radicle is united to the oxygen atoms 



— C — . and not to the carbon atoms, which determine the isomerism. 



j 'BOH ït is doubtful if it may be applied to cistrans non-saturated 



— C — acids; if the formation of complex metallic compounds must 



I be represented by the following formula: 



Hg-0 

 or anyhow, if it is to be supposed that the double bond is attacked, we can rely 

 upon this method as little as on all others, by which the bearers of the isomerism 

 are affected (see below). 



Now the contrast between the two groupings H and COOH, which 

 are bound to the very simple skeleton of maleic and fumaric acid, 

 is exceptional; besides, the two carboxyl groups may enhance each 

 other's acidity or react with one another under anhydride formation. 

 Also a mutual repulsion of the COOH groups is to be anticipated, 

 from which could be deduced that fumaric acid is more stable than 

 maleic acid. 



The particularly simple structure of these acids, by which the 

 carboxyl groups take the leading position, has rendered the above- 

 mentioned considerations so successful in determining the configurations. 



As soon as tiiis reciprocal influence fails or the structure becomes 

 much less simple, we have no longer any certainty. 



We will examine «- and wo-crotonic acid in this respect: 



1. The dissociation constants are: «-crotonic acid 2 . JO"''. 



iao- ,, ,, 3,6 . JO— ^. 



By comparing propionic with acetic acid it could be deduced that 

 a metliyl group weakens the acidity, so tliat in «-crotonic acid the 

 methyl group must lie on tiie same side as the acid group; in fact 

 acrylic acid {k = b,Q .10-^) is dissociated in a higher degree tiian 



6 



Proceedings Royal Acad. Amsterdam. Vol. XXI. 



