74 REPORT OF SCHIMMEL & Co. OCTOBER 1915. 
It is very important to decide whether the action of ozone may cause a change 
in the position of the double linking. The cases where such a transposition is possible 
are not very numerous. The groups ae PC: C— and CH eC , as occurring for 
instance in rhodinal (IV) and citronellal (1), come chiefly into consideration. The results 
afforded by the action of aqueous potassium permanganate ') on citronellal have but 
one meaning, as only derivatives of rhodinal are obtained. Intermediate products can 
only be isolated with the aid of potassium permanganate in acetone. Through the 
action of ozone, however, products are formed which indicate that the commercial 
citronellal is a mixture’). A content of 40 p.c. of rhodinal can be deduced from the | 
quantities of acetone or acetone peroxide and 6-methyladipic acid, as the only decom- 
position products that could be isolated at the beginning. 
Harries continued these investigations together with G. Wagner’). He did not sila 
the ozone to act direct on the aldehyde, as .before, but on its dimethylacetal. But 
even then the decomposition did not lead up to the desired result, as principally only 
acetone peroxide and -methyladipic acid could be traced, in addition to a new body, 
a keto-aldehyde. It became evident from this experiment that in the decomposition of 
the ozonide of citronellaldimethylacetal by water the acetal groups were split off, so 
that the aim of protecting the aldehyde group against alterations by acetalizing it had 
not been gained. 
Later on, Harries again tried the problem in company with F. Comberg*). As an 
indifferent product, the peroxide of methyloctanonal (Il) was isolated; on being heated 
with potash lye it passes into ethanoyl-4-methylcyclohexene-1 (III), a cyclic derivative. 
It was easy to determine the constitution of this compound, as it proved to be identical 
with the ketone tetrahydro-p-acetotoluene, which Wallach*) had prepared synthetically 
some years before. This tetrahydro-p-acetotoluene can only have formed -from the 
peroxide of the ketoaldehyde (II), which must resuit from the genuine citronellal by 
oxidation with ozone. 
Whilst in this way the presence of genuine citronellal in the commercial product 
is proved, on the other hand a series of compounds may be obtained which indicate 
an admixture of rhodinal. In addition to the 6-methyladipic acid previously observed, 
the authors obtained its half-aldehyde and acetone peroxide, as well as a new acid, 
which they consider to be 5-methyleyclopentenecarboxylic acid (V). 
It results from these experiments that commercial citronellal is not uniform, with 
which, however, dogs not agree the fact that it yields a semicarbazone, described 
already by Tiemann and Schmidt, and which uniformly melts at 84°. In consequence, 
one tried to split the semicarbazone into its components by crystallization, but no 
alteration of the melting point could be stated. However, if the semicarbazone was 
treated with ozone, an ozonide formed which, on decomposition, yielded two semi- 
carbazones. In addition, formic acid, carbon dioxide and acetone peroxide result. The 
two semicarbazones were identified as methyloctanonal monosemicarbazone (VI), which_ 
corresponds to normal citronellal, and as the semicarbazone of the half-aldehyde of” : 
the y-methyladipic acid, which is derived from rhodinal (VII). The proportion calculated 
is therefore 60°/o of citronellal to 40 °/o of rhodinal. It might be concluded from the 
experiment in question that the semicarbazone is an individual substance in which a 
shifting of the double linking is produced by ozonisation. Harries does not consider 
1) Tiemann and Schmidt, Berl. Berichte 29 (1896), 903; Report October 1896, 88. — 7?) Harries and 
Himmelmann, Berl. Berichte 41 (1908), 2187; Report October 1908, 185. — #*) Inaugural-dissertation, Kiel 1913. 
— 4) Inaugural-dissertation, Kiel 1914. — 5) Liebigs Annalen 360 (1908), 54. 
