i8 SCIENCE PROGRESS 



The present paper deals with the first acid and indicates that 

 ring formation takes place comparatively readily. On the 

 other hand, the investigation of bromo-adipic acid, of which 

 the author promises details in a subsequent paper, showed a 

 marked difference in the ease of formation of the tetramethylene 

 ring. The author's results, therefore, are so far in accordance 

 with the thermo-chemical results and his own hypothesis. 



The Constitution of Pinacyanol. — Although pinacyanol 

 (diethylcarbocyanine iodide), prepared by the action of alkali 

 and formaldehyde on a hot alcoholic solution of a mixture of 

 quinoline ethiodide and quinaldine ethiodide, was patented as 

 a photographic sensitiser as early as 1905, its structure has been 

 a matter of doubt. 



Two formulae have been suggested, [XI] by O. Fischer 

 and [XII] by Wise, Adam, Stewart, and Lund. Mills and 

 Hamer {Trans. Chem. Soc, 1920, 117, 1550) prefer, however, 

 that pinacyanol should be represented as [XIII]. 



C : CH \JkJ EtN 



NEt lEtN 

 [XI] 



CH • CH : CH 



NEt lEtN 



, [XIII] 



There seems to be no doubt that pinacyanol contains but 

 one monovalent acid radicle, and from careful analyses of the 

 iodide and bromide the molecular weight has been shown to 

 be 479 ± I. This rules out Fischer's formula. 



Although better yields of the pinacyanol are obtained when 

 a mixture of quinaldine alkyl iodide and quinoline alkyl iodide 

 is used rather than the quinaldine compound alone, the quino- 

 line alkyl iodide plays no direct part in the condensation. 

 This was discovered by Fischer, and confirmed by the present 

 authors by showing that by the action of alkali and formal- 

 dehyde on a mixture of p-toluquinaldine ethiodide and 

 quinoline ethiodide the dye produce was homogeneous, and 

 consisted of i : i'-diethyl-6 : 6'- dimethylcarbocyanine iodide. 

 The untenability of the formula [XII] is shown by the oxidation 

 of pinacyanol. Oxidation of the bromide with dilute nitric 

 acid yields quinaldinic acid ethyl nitrate [XIV] in amounts 

 corresponding to about 90 per cent, of the calculated quantity 



