Notes on scientific research. 113 



and 263.0 = 84.3 per cent, citronellol) were separated into 6 and 5 fractions respec- 

 tively. It was then found that there had been formed on the average, in formylation: — 



With the With the 



100 per cent, acid 85 per cent, acid 



Terpenes and citronellol . . . . 1 per cent. 3 per cent. 



Citronellyl formiate ...... 20 „ „ 30 „ 



Citronellyl glycol monoformiate 1 ) 35 „ „ 40 „ „ 



Citronellyl glycol diformiate 2 ) . 29 „ „ 12 „ „ 



Residue, polymerized 15 „ „ 15 „ „ 



By the same treatment, citronellyl formiate gave rise to a product of the following 

 composition: — Unchanged citronellyl formiate, 35 per cent; citronellyl glycol di- 

 formiate, 50 per cent; polymerized residue, 15 per cent 



Hence the exceedingly high values obtained by determining pure citronellol by 

 the formylation method 3 ), are due to the formation of citronellyl glycol diformiate. 



Pfau concludes, as result of his work, that the formylation process for estimating 

 citronellol furnishes irregular values and affords no clue as to the true content of 

 citronellol. This is nothing novel, as we always have pointed out 4 ) that this process 

 is not an exact one, although it may be satisfactory in practical work. 



Equally, C. T. Bennett 5 ) tested the formylation method by varying the conditions 

 and came to the conclusion that the method is unreliable for the determination either 

 of citronellol or citronellal. 



With regard to these two articles by Pfau and Bennett, W. H. Simmons 6 ) is of 

 opinion that both workers confirm his own conclusions on the subject, but apparently 

 ignore what he considered to be most important, viz., that the method is a very useful one 

 for comparing the relative proportions of citronellol in different geranium oils, as was 

 suggested by him in 1913. Subsequent experience with a very large number of samples 

 of African and Bourbon geranium oils fully confirmed the view which Simmons then 

 expressed. It is very important to adhere strictly to one and the same direction and 

 to heat for exactly 60 minutes on a boiling water bath, and not on a sand bath. 



For the estimation of terpin hydrate in elixir of terpin hydrate Nat. Form. 7 ), 

 A. G. Murray 8 ) gives the following directions: — The elixir is mixed with a solution 

 of 20 gr. common salt in 100 ccm. of water until the alcohol content is from 10 to 

 15 per cent, by volume. The mixture is shaken out with four portions, one-fourth 

 volume each, of chloroform containing 5 to 7 per cent, by volume alcohol. Each 

 portion of the solvent is washed successively with 5 ccm. salt solution, then filtered 

 into a tared beaker, and evaporated by aid of a blast, avoiding application of heat. 

 (On evaporating an alcoholic solution of terpin on the water bath, about 10 per cent, 

 of the terpin hydrate were found to volatilize.) The residual pure terpin hydrate is 

 dried at the open air and then weighed. The error was only -f- 0.8 per cent. 



The acetylation of both hydroxyl groups in terpin may be carried out by boiling the 

 glycol for three hours with eight times its weight of acetic acid and five times its weight 



a ) d l5O 0.9651; a -f1°46'; n D15 ol.4488. — 2 ) d 15 o 0.9976; a D -f-l°33'; n D15 ol.4425. — 3 ) Cf. Eeport 

 October 1913, 64. — *) Gildemeister and Hoffmann, The Volatile Oils, 2 nd ed., vol. I, p. 600. — 6 ) Perfum. 

 Beeord 12 (1921), 351. — 8 ) ITddem 398. — 7 ) It contains, in addition to about 1 grain (0.175 gr.) of terpin 

 hydrate, tincture of sweet orange peel, spirit of bitter almond, alcohol, glycerin, and syrup in 1 fl. dr. — 

 *) Journ. Amer. pharm. Ass. 10 (1921), 440. 



