— I2 4 — 



The authors have recently proved 1 ) that methyl ethyl ketone is also 

 split up, in an analogous manner, into acetic acid and ethane, and 

 not (as might have been equally possible) into propionic acid and 

 methane. Laevulic acid, that is to say a ketonic acid, should in an 

 analogous manner be split up into a mixture of acetic and propionic 

 acids, but only the acids propionic acid and formic acid, formed 

 according to the reaction 



CH 3 .CO.CH 2 .CH 2 .COOH-> CH 3 OH+HCOOH-f CH 3 .CH 2 .COOH 



could be detected, and the formation of methyl alcohol was rendered 

 probable. At any rate, only traces of methyl alcohol could be detected 

 by oxidation to formaldehyde and by treatment of the repeatedly 

 distilled first portions with hydrogen iodide according to Zeis el. When 

 the silver salts of the acids were produced, the precipitate became 

 deeply blackened during recrystallisation from water, which pointed 

 to the presence of formic acid. The purified silver salts gave numbers 

 which agreed well with propionic acid. In alcoholic solution the laevulic 

 acid was partly reduced to ^-hydroxy valeric acid which occurred 

 as lactone; another portion was esterified. From the alcohol there 

 was formed partly acetaldehyde, corresponding to the reduction of the 

 acid. A remarkable fact was the splitting up of menthone, with which, 

 as a cyclic body, no splitting up of the molecule could be expected, 

 but only an opening of the ring. Aqueous-alcoholic solutions of 

 menthone (altogether 260 g. ketone, 975 cc. alcohol, 520 water) which 

 had been exposed to the light during the summer and autumn months, 

 separated off a layer of oil on the surface, and showed an acid reaction. 

 The solution neutralised with soda, when extracted, yielded besides 

 unchanged menthone, an aldehyde which imparted to the solution 

 a peculiar odour, and which could be separated off from the menthone 

 with bisulphite. From 160 g. d-menthone 11,5 g. aldehyde of the 

 b. p. 193 to 1 97 were obtained; after further purification through 

 the bisulphite compound, the body boiled absolutely constantly 

 at 1 95 . 



The combustion led to the formula C 10 H 18 O, that is to say, 

 an isomeride of citronellal; but the two aldehydes were not identical, 

 as the new aldehyde yielded a semicarbazone of the m. p. 88 to 89 , 

 and citronellal one of the m. p. 81 to 82 °. The hydroxamic acid of 

 the aldehyde melted at 108 to 109 . Ciamician and Silber assume 

 that the body is the mentho - citronellal produced by Wallach from 

 meiitho-citronellol (obtained by splitting up menthone oxime), whose 

 semicarbazone also melts at 89 . The conversion would be accomp- 

 lished as follows: — 



Berl. Berichte 40 (1907), 2415. 



