— 27 



foot by de Jong 1 ) with this object have been greatly facilitated by 

 the information furnished by Bacon 2 ). The latter has shown that 

 most first-class Manila ylang ylang oils have a sap. no. of above ioo, 

 while the second grade, of which the value equals that of cananga 

 oil, always has a sap. no. below ioo. De Jong found that fractions 

 of oils distilled in various factories at Serang possessed the following 

 constants: — 



Brand 



«D 



Sap. no. 



Sap. 



no. after acetylation 



A. 



-2 3 °o' 



27,1 





76,7 



B. 



-28° 30' 



24,1 





17fi 



i. 



— 24 10' 



29,1 





— 



2. 



- 29° 45' 



33»o 





— 



3- 



-35°o' 



32,6 





— 



As the ester and alcohol content of these fractions was much 

 lower than that of the ylang ylang oil itself (which may possibly be 

 due to the insufficient cooling during distillation), de Jong distilled some 

 flowers himself, which he obtained from the neighbourhood of Tjibin- 

 veng and Tjiterep, near Depok. Partly green and immature flowers 

 yielded the following fractions: — 



1. «d — 2i°32 / ; sap.no. 17,5 



2. «d —39° ; sa P- no - 23,8 

 3- °d —35° 4o'; sap. no. 23,0 



Red, over-mature flowers also, either distilled whole or after being 

 pounded, produced only fractions with a rotatory power ranging between 

 «d — I2 ° 56' and a D — -35°o', and a sap. no. between 9 and 17. 



Oil from Batavia flowers gave fractions of «p — 20 37' to — 3 8° o' 

 and sap. no. 33,6 to 37,3. This oil is better than that from the Depok 

 district, and resembles rather the product obtained in Serang. 



According to de Jong, the difference in the constitution of the 

 oil depends upon the site of the trees and upon climatic conditions, 

 but it may also be due to minor botanical differences in the trees. 

 Efforts are to be made to obtain seed and seedlings of Cananga 

 odorata from the Philippines, and to acclimatise them in Java, in order 

 that the causes of the difference in the constitution of the Philippine 

 and Java oils may hereafter be made clear. 



De Jong further lays special stress upon the fact that the vola- 

 tility of the esters contained in the oils under examination is not 

 greater than that of their other constituents, because the sapon. numbers 

 of the successive fractions show very little variation. 



x ) Milit. Tijdschrift 1908, 1. Batavia. From a copy kindly sent to us. 

 2 ) Comp. Report November 1908, 128. 



