— 154 — 



The object of our investigations was in the first place to determine 

 whether hydroxylamine hydrochloride is suitable for the quantitative 

 determination of aldehydes, especially of citral. For this purpose, 

 aldehydes of known constitution were taken, viz., 



i. Several samples of pure citral prepared by ourselves, including 

 citral from oil of lemon, the content of which by the use of neutral 

 sodium sulphite was ascertained at 99%. 



2. Solutions of acetic aldehyde, of which the aldehyde -content 

 was ascertained at 50 /o by titration with sodium bisulphite and 

 iodine solution. 



3. Freshly distilled chlorine-free benzaldehyde, b.p. 178,5° (751 mm.) 

 of which the purity was also shown by the clear solution with water 

 made by its bisulphite compound. 



The values determined according to Bennett were, for 

 Citral 88 to 93 % in lieu of 99%, diff. about 9% fcompared with 



Acetaldehyde 44 „ 45% „ „ „ 50%, „ „ 1 1 % { 100 per cent. 

 Benzaldehyde 92 ,,93 % „ „ ,,100%, „ „ 8% l * 



The figures found were therefore about 10% too low. The citral- 

 content of various terpenes from oil of lemon and oils of lemon was 

 also determined, and known quantities of pure citral admixed to the 

 samples. The citral- content was then ascertained afresh, and again 

 found to fall about io°/ below the values calculated. 



Lemon-oil terpenes: 



0,25 °/ citral -content, +5% citral S ave 4> 8l > /o = 4' 6 % instead of 5 % 

 0,6 % » +5% » » 5' 2 °/o = 4.6J7o » » 5% 



I » 2 /o » t5 /o »» »» 5»5 /o — 4»3 /o »» »» 5 /o 



Lemon oil: 



3,9% citral -content, + 3 % citral gave 6,5 °/ = 2,6 °/ instead of 3 % 



4 °/ „ +4,4% » » 7,8 /o = 3,8 /o „ „ 4,4% 



Although it was made clear beyond doubt that Bennett has improved 

 Walther's method, the above experiments show that even the improved 

 method does not furnish exact results, as the values found averaged 

 about 10% too low, calculated for citral present. The error probably 

 does not matter very much in the case of oils of lemon, with their 

 comparatively low citral-content, and the determination may therefore 

 suffice for practical purposes, especially as it is easily carried out. But 

 where oils of high percentage (lemongrass oil) are concerned, the error 

 will naturally be much greater, and for such oils, therefore, Bennett's 

 method does not appear to be admissible. 



In our last Report 1 ) we discussed a process described by A. G. Wood- 

 man and E. F. Lyford for the colorimetric determination of benz- 



x ) Report April 1909, in. 



