— 202 — 



alcoholic potash -liquor only one half had been attacked. Free a-cyclo- 

 geranic acid melted at 106°, and could only be esterified with difficulty, 

 but by heating the acid in an autoclave to 120° with sodium ethylate 

 and ethyl bromide Bouveault succeeded in preparing the ethyl ester: 

 b. p. 101 to 102° (10 mm.). By treating a-cyclogeran\c acid with penta- 

 chloride of phosphorus he obtained the chloride, from which upon 

 saturation with gaseous ammonia in ethereal solution, a-q/dogeranic acid 

 amide (m. p. 120 to 121 °) was formed, which makes it very probable that 

 the amide melting at 202°, to which reference has been made above, is 

 that of ft-cyclogevanic acid. By reducing with sodium and alcohol, he 

 obtained from the ethyl cycloge rani ate, a-q/cfogeraniol 1 ), a colourless 

 liquid of a pleasant odour, boiling at 213 to 217°, of which the ethyl 

 ether boils at 115° (20 mm.); m. p. of the phenylurethane 75°. For 

 identification-purposes an attempt was made to prepare the cyclogevanyl 

 ester of pyruvic acid, but this reaction, instead of yielding the looked-for 

 ester, resulted in the formation of a body possessing a similar empiric 

 composition, but a different chemical constitution. In all probability ring- 

 formation and the splitting-off of water had occurred, and resulted in the 

 forming of a compound with the following formula: — 



CH 3 



CH 2 



CO 



H 2 C 

 H 2 C 



C— O O 



CH-CH^ 

 CH 3 



CH 8 

 CH 2 



When a-cycloger anic amide (m. p. 121°) was reduced with sodium and 

 amyl alcohol, a-cyclogeranio\ was also formed, buth when ethyl instead of 

 amyl alcohol was used no reduction took place. 



Commercial q/cfogeraniol, prepared by the isomerisation of geranyl 

 acetate, when treated with pyruvic acid also yields a true ester of pyruvic 

 acid, in addition to the compound melting at 114° which has been referred 

 to above, and from this ester of pyruvic acid the corresponding alcohol could 

 be regenerated, showing that commercial cyclogevamol is not a uniform body. 



In a preliminary communication, Gardner, Perkin and Watson 2 ) de- 

 scribe a" series of acids which they have prepared by treating cyclic ketones 

 in ethereal solution with sodium amide and carbon dioxide. Among others 



*) Bull. Soc. chim. IV. 7 (1910), 354. 



2 ) Proceed, chem. Soc. 26 (1910), 136; for details see Journ. chem. Soc. 97 (1910), 354. 



