I - 41 _ 



ellal is carried out in one way or in another, we think it necessary, in 

 order that errors may be avoided, to point out that with the original oil 

 the acetylation method is a matter of considerable importance. With such 

 oil it is absolutely necessary to acetylate with the addition of sodium 

 acetate; the best way of doing so being by the method frequently quoted 

 by us for citronella oil. This consists in keeping each 10 g. of oil and 

 acetic anhydride with 2 g. melted sodium acetate to boil uniformly on the 

 sand-bath for 2 hours. 



We have tried the oximation method in the first place upon several 

 citronellal-samples of our own manufacture, and have found that after 

 oximation and after most careful purification the proportion of acetylisable 

 constituents in these samples was considerably in excess of 2,5 p. c. 

 Whether this is the fault of the method or whether the sensitive citron- 

 ellal is partly readily converted into another compound, cannot be said 

 for certain. The characters of a citronellal which had been kept for some 

 time in a closed bottle were as follow: b. p. 65 to 66° (5 mm.), di 5 o 0,8659> 

 « D +9°34', n D20O 1,44962, sol. in 3,5 vols. a. m. 70 p. c. alcohol. After 

 acetylation we found 100,6 p. c. Ci Hi 8 O and after oximation 22 p. c, 

 showing a citronellal-content of 100,6 — 22 ^78,6 p. c. We then tried to 

 separate the citronellal from the ^opulegol (which is more soluble in 

 water), by fractionation and by repeated steam-distillation, but we only 

 succeeded imperfectly in doing so. 



As it was impossible to obtain a pure product by this means, the 

 aldehyde was regenerated from the carefully purified bisulphite compound 

 and was purified by steam-distillation : di 5 o 0,8562, « D +10°47', n D20 o 1,44561, 

 sol. in 5 vols. a. m. 70 p. c. alcohol. The viscous and almost odourless 

 oxime obtained from this pure citronellal gave di 5 o 0,9083; « D — 7° 40'; 

 no2oo 1,47151; sol. in 4 vols. a.m. 60 p. c. alcohol; sap. v. 1,5. After one 

 hour's acetylation of the oxime we found 13,3 p. c. Ci Hi 8 O, or conse- 

 quently 86,7 p. c. citronellal; after l 1 /* hours heating with acetic anhydride 

 according to Dupont and Labaune, 88,5 p. c. In order to discover whether 

 possibly the oxime still contained aldehydic constituents it was oximated 

 afresh. Subsequent acetylation once more gave only 87 p. c. citronellal. 



We have already expressed the surmise that the differences in the 



estimation of citronellal may possibly be due to its instability which, as 



^experiments showed, manifests itself in its constants after a short time. 



A sample gave di 5 o 0,8598, « D +9° 38'. After 4 weeks the constants 

 of the sample, which had been kept in a half-filled bottle, were di 5 o 0,8619, 

 "d+9°50\ 



A sample recovered from the bisulphite compound gave at first di 50 

 0,8574, « D + 10°3f; after 6 weeks it gave d 15O 0,8730, « D + 10 o 50\ 



In addition to our experiments with pure citronellal we have also 

 tried the method with mixtures of known aldehyde content. These mix- 

 tures have also been employed in the other methods mentioned below 



