— 151 — 



1 to 2 gr. chlorine-free hydrate of lime are stirred up with a glass 

 rod in a small porcelain dish with 10 to 15 drops benzaldehyde, the 

 mixture is then covered with a thin layer of hydrate of lime and slowly 

 brought to a red heat. After cooling the contents of the dish are 

 poured into a beaker, 5 to 6 cc. water are carefully added, and the 

 mixture is faintly acidulated with nitric acid. The solution is then 

 filtered through chlorine-free filtering paper or glass-wool and is tested 

 for chlorine with silver nitrate. 



By this method Heyl was able to trace clearly so small a proportion 

 of chlorine as 1 drop of monochloro-benzene in 50 gr. benzaldehyde. 



Rupp proposes for the same purpose the copper oxide test (the so- 

 called Beilstein test) for which he gives the following modus operandi: — 



A copper wire, or, better still, a strip of copper wire-netting 0,5 cm. 

 in width (about 1 mm. mesh) is twisted at one end into a narrow spiral 

 shape, forming a roll about the diameter of a large pea. This roll is 

 repeatedly drawn through a non-luminous flame in order to oxidise the 

 surface of the copper and to remove any yellow or green coloration. 

 When cooled, the roll is dipped into the benzaldehyde under exam- 

 ination and is then exposed to the flame for a moment in order to in- 

 flame the benzaldehyde. When the benzaldehyde is burnt off (not in 

 the flame) the roll is again applied to the top of the flame. If there 

 is a green inflammation, chlorine is present. 



A roll of netting absorbs about 0,3 gr. and a wire-spiral about 0,1 gr. 

 of the substance to be tested. Rupp found that when a roll of wire netting 

 was used, so small a mixture as 1 drop benzylchloride in 50 gr. benz- 

 aldehyde gave a clearly perceptible reaction, whereas with copper-wire a 

 mixture of 50 gr. benzaldehyde and 2 drops of benzylchloride was needed 

 to produce such a reaction. 



We ourselves have also suggested both these methods for the de- 

 tection of chlorine in essential oils and preparations 1 ), but nothwithstanding 

 this we prefer the combustion method for the purpose here under dis- 

 cussion, that method having been specially worked out for the testing of 

 bitter almond oil for chlorine. Our experience shows it to be superior 

 to the lime-test (although in itself that test is a sensitive one), inasmuch 

 as it can be carried out more rapidly, especially when a series of oils is 

 to be tested. And where minute proportions of chlorine are concerned, 

 tie copper test is decidedly inferior to both the combustion and the lime- 

 methods, not only in respect of sensitiveness, but especially in respect 

 of certainty of action. Hence the copper test can only be recommended 

 under reserve and for the purposes of a preliminary examination. 



x ) Gildemeister and Hoffmann, Die atherischen die, 2 nd Ed., Vol. I, p. 630. 



