46 Report of Schimmel 8 Co. April 1913. 



9° sol.pt. according to the phenylhydrazine method, to be 100 p. c. calc. for methyl- 

 nonylketone. In reality this value was of course too high, owing to the presence in 

 the oil of methylheptylketone as well as methylnonylketone. 



From the analysis of a citronella oil sample recently submitted to us, we believe 

 ourselves to be justified in concluding that certain kinds of motor-spirit (automobile- 

 benzine) have lately been used as adulterants of Ceylon citronella oil. The sample in 

 question showed the following constants: d. 25 o 0,8873, « D — 11° 12', so-called total 

 geraniol 55,6 p. c; sol. in its own vol. of 80 p. c. alcohol; upon the addition of 5 vols, 

 of the solvent the solution becomes almost turbid. The sample barely answered 

 Schimmel's test, and failed to answer Schimmel's "raised test." 



Although the percentage of so-called total geraniol must be described as tolerably 

 sufficient, the oil is obviously suspect on account of its sp. gr., which is much too 

 low, and because it is doubtful whether it can or cannot be said to answer Schimmel's 

 solubility test. When the sample was subjected to further examination, it was found 

 that this abnormality was. clearly due to sophistication, and that the oil contained 

 about 10 p. c. of a fraction boiling between 80 and 160°, whereas a check test shewed 

 that of a sample of normal oil practically nothing at all passed over below 160°. 



Unfortunately, the sample sent to us was insufficient for a thorough examination 

 of the fraction in question; but it would appear that it consisted of low-boiling petroleum 

 hydrocarbons. This suspicion was confirmed because experiments showed that it is 

 quite possible for a citronella oil of good quality still to answer the solubility-test 

 after an addition of 10 p. c. or even more of benzine. But apart from this the adulteration 

 could not have escaped the notice of a careful observer, inasmuch as it can be imme- 

 diately detected (as was also the case with the sample in question), from the reduced 

 sp. gr. ; a constant which in normal Ceylon citronella oil ranges from 0,900 to 0,920. 

 It will therefore be advisable in future examinations to take special notice of this 

 constant, but in addition to this we would again call attention to our "raised test" 

 (testing the solubility of a sample after adding 5 p. c. petroleum), as this test would 

 in most cases, we think, reveal adulterations of the character described. Presumably 

 the method recommended by Dodge 1 ), consisting of oxidation with potassium perman- 

 ganate, would also be serviceable in detecting an addition of the above nature. In 

 applying this test the adulterant, which consists entirely of hydrocarbons, would remain 

 unattacked. 



We hope to be able to give further particulars on this subject as soon as we 

 have examined a larger quantity of oil sophisticated in this manner. 



On the detection of petroleum in citronella oil see p. 129. 



On a Burmese citronella oil see p. 71. 



Clove Oil. Since the publication of our last Report the prices of clove oil have 

 been repeatedly advanced, and it appears to us that the end of the upward movement 

 has not yet been reached. Events have shown that our estimate of the yield of the 

 summer crop was still too high, and it is now certain that the winter-crop also has 

 not yielded more than 30 000, or at the most 40 000 bales, while the world's annual 

 requirements of cloves amount to 120 000 bales. It is therefore not surprising that the 



l ) Comp. p. 62 of the present Beport. 



