558 SCIENCE PROGRESS 



points are determined in some suitable solvent for the two forms 

 separately and then for the two forms together. If polymor- 

 phic, the mixture of the two forms will give a point lying between 

 those of the separate forms ; but if isomeric (or tautomeric) 

 the point given by the mixture will lie below that of the more 

 soluble form, the depression of freezing-point by the two to- 

 gether approximating to the sum of the individual depressions. 



The method has the great advantage that as it takes only a 

 few minutes to carry out at a low temperature, the probability 

 of the two forms (if indeed they are not polymorphic) coming 

 to tautomeric equilibrium in the course of the experiment is 

 greatly diminished. 



It seems likely that many supposed cases of isomerism will 

 prove on close examination to be instances of polymorphism. 

 A recent example of this is worth recalling. 



According to current views on configuration two cinnamic 

 acids of the formula C 6 H 6 CH : CH . COOH should exist 



CeHo — C — H CgHs — C — H 



II 11 



COOH— C— H H— C— COOH 



Cis. Trans. 



but at least four have been described. 



The best known is the acid whose aldehyde occurs in oil of 

 cinnamon and which is found in various balsams. It is gener- 

 ally made by a synthetic method discovered by Sir William 

 Perkin : benzaldehyde is heated with sodium acetate in presence 

 of acetic anhydride. This acid crystallises in needles or prisms, 

 melts at 133 , and boils at 300 . 



Three other acids of identical composition and chemical 

 properties melting at 68°, 5 8°, and 42 ° respectively have been 

 obtained from coca leaves and by various other methods, in- 

 cluding the partial reduction of phenylpropiolic acid. They 

 must therefore have the cis. configuration. It follows that 

 ordinary cinnamic acid has the trans, configuration. 



Many fruitless attempts have been made to devise formulae 

 to explain the existence of these three low melting acids. They 

 have recently been shown to be simply polymorphic forms of 

 the cis. acid. 



By melting either of the three acids identical fusions are 

 obtained and the introduction of a crystal fragment of any 



