SCIENTIFIC NOTES ON ESSENTIAL OILS. 13 
more geraniol ‘than citronellal. Whereas with high-class Java oils, containing 92 to 
94 p.c. of so-called total geraniol, geraniol proper amounted to about 50 p.c. and 
citronellal to perhaps 44 p.c., these inferior oils showed only 79,5 to 82,5 p.c. of 
so-called total geraniol, whereof about 50 to 58 p.c. was geraniol proper and from 
22 to 32 p.c. citronellal; moreover, the specific gravity of these oils was higher 
(0,899 to 0,901) than in the case of good oils (0,884 to 0,897)*). The author of the article 
in question seems to be in some doubt as to the causes of this anomaly. He seems to 
think that it might possibly be due to some Ceylon citronella oil having been added 
to the Java oil; or to the raw material, the grass used having perhaps been some 
hybridized plant, closely related to the Ceylon variety, or that some differences in the 
cultivation reacted on the quality of the grass; he also touches upon the possibility 
of overheating in the course of distillation. His observations lead him to the con- 
clusion that, apart from the total geraniol percentage, it would be advisable to value 
the oil also according to its percentage of citronellal, by establishing a mininum limit 
for this latter body. 
To obviate misunderstandings we would add that the geraniol percentage was 
ascertained according to the method of Dupont and Labaune. In our Report on this 
method?) we had pointed out that not only the geraniol but all the alcohols present 
in the oil, either free or in the form of ester, are ascertained thereby, thus offering 
more especially indications as to the quantity of citronellal, which can be deduced 
from the difference between the so-called total geraniol and the alcoholic parts. For the 
test of geraniol proper the only suitable process is our phthalic anhydride test?). As. 
mentioned in our previous Report (page 44) by means of this test we found from 
26 to 44,4 p.c. of geranio!l in Java citronella oils. The citronellal percentage, as 
established by Dupont and Labaune’s method, amounted to 23,4 to from 50,1 p.c. We 
stated at the time that oils with a high citronellal percentage were almost invariably 
of especially fine odour, and of a low specific gravity. 
Of oxides of the terpene group, represented in essential oils, only cineol, Ci.HiO, 
and calameone, C;;H2,,O2., were known so far, whereas carlina oxide, CisHi90, must 
probably be looked upon as a phenyl-1-e-furyl-3-allene. A body, CooHs,O, which was 
proved by K. E. Spornitz*) in the high-boiling (b. p. 190 to 200° at 15 mm.) parts of Java 
citronella oil and which was called dicitronella oxide, is to be considered as an oxide 
of the diterpene group. It was distilled over sodium, yielding a colourless liquid which 
boiled from 182 to 183° (12 mm.): d2>0,9199; %) —4°; np 1,49179; mol. refr. found 91,43, 
calculated for CsoH:10/; 90,85. When boiled with acetic anhydride and sodium acetate 
this body formed no ester, neither did it split off water with concentrated formic acid, 
nor change when treated with sodium and alcohol. Elementary analysis as well as deter- 
mination of molecular weight in benzene solution, after Beckmann, agreed with the 
formula C2.H;,0. By the introduction under pressure, in the presence of platinum 
black, of hydrogen into the ethereal solution or into the glacial acetic solution, 
dicitronella oxide may easily be reduced to tetrahydrodicitronella oxide, CooH330; b. p. 180 
to 185° (11,5 mm.); d=>0,9001; np 1,47457. By introducing dried hydrochloric acid 
gas into the ethereal solution of dicitronella oxide, a monohydrochloride, m. p. 107,5°, 
1) The rotation of the inferior oils is given as —1 to —2°, of the normal oils as +1°. This might lead 
to the belief, that dextro-rotation is a characteristic of Java citronella oils. This is not the case; on the con- 
trary Java citronella oils are mostly laevo-rotatory (generally not exceeding — 3°), whereas dextro-rotation is the 
exception, and not the rule. See Gildemeister and Hoffmann, The Volatile Oils, 2.4 ed., vol. H, p. 241. — 
*) Report October 1912, 40. — *) Ibidem 43. — +4) Berl. Berichte 47 (1914), 2478. 
