104 BACON. 



point, 106° to l«8°at 7(iO millimeters: specific f'lavitv'- =0.8354; X' -_ = 1 .4058 ; 

 A—- =91.5. Tlii.s terpene oil gave an abundant precipitate of pliellandrene nitrite 

 and was considered to be j3 pliellandrene. 



LIMONEXE. 



The various samples of terpenes from ele»n which were considered 

 to be pure (i-limonene were united and after three distillations over 

 sodium gave an oil boiling between 176° to 177°.4 (thermometer whollj- in 



30° 30° 30° 



vapor) ; with a specific gravity, -^=0.8350 ;.N'-j^=1.4G70; A^- = llo.7. 



This oil yielded limonene tetrabromide melting at 104°. Bepeated dis- 

 tillations over sodium always gave the same range in the boiling point 

 and no change in the optical activity, and therefore, I consider this to 

 be a very pure sample of limonene. 



I have obtained more satisfactory results in preparing the hydi'ochlorides 

 and hydrobromides of terpenes, by using petroleum ether to dilute the terpene, 

 instead of carbon bisulphide as recommended by Wallach. Petroleum ether has 

 the advantage of permitting the terpene and the solvent to be distilled over 

 sodium directly before the dry hydrogen halide is run into the mixture and absolute 

 certainty that everything is anhydrous is assured. Limonene hydrochloride pre- 

 pared with petroleum ether as a solvent is a much purer product than when 

 carbon bisulphide or any other solvents are used; the body made in this way 

 boils within 2° or 3° on the first distillation in vacuo, while with carbon bisulphide 

 a range of 10° to 15° is common and several fractionations are necessary to obtain 

 a pure product. 



I have also been able to prepare pure limonene hydrobromide and phellandrene 

 hydrochloride by this method; tlie properties of these bodies are not recorded in 

 the literature.'" 



In a previous article ^' in this series, I have shown that limonene 

 hydrochloride reacts with magnesium according to the Grignard reaction, 

 the product of the decomposition of the hydro-limoiiene magnesium 

 chloride with water being a dihydroterpene CmHig. This dihydroterpene 

 instead of the expected earbinol was also obtained with benzaldehyde, 

 the benzaldehyde behaving as if it contained a h3'droxyl group. A con- 

 tinuation of this line of research has sho^^^l that while terpene halides in 

 general easily react with magnesium in the presence of ether, the addi- 

 tion products so formed do not readily enter into reaction with other 



"A uniform, easily controlled stream of dry hydrogen bromide may be obtained 

 by the following method: Red phosphorus is covered with a layer of completely 

 saturated, aqueous hydrobromic acid, and a mixture of 3 parts bromine and 1 part 

 saturated hydrobromic acid is dropped slowly into this mixture. The gas is 

 evolved quietly and regularly and after passing two wash bottles containing red 

 phosphorus, suspended in saturated hydrobromic acid, is dried by means of a 

 long tube of granulated calcium chloride. There is practically no contamination 

 with hydrochloric acid by this inethod, although calcium bromide would be better 

 for drying purposes. 



" This Journal, Sec. A. (1908), 3, 49. 



