— 120 — 



by the method described below will be found to contain chlorine. This 

 we have proved absolutely by comparative tests. Herzog is right in 

 criticising the wording of the Pharmacopoeia-method, which should 

 have been much more explicit, for as it is at present it may lead first 

 to misunderstandings and subsequently to such unsatisfactory results as 

 have been obtained by Herzog. But when this test is carried out in 

 a proper manner it is so extraordinarily delicate that it will indicate 

 even the smallest traces of chlorine-products. In the course of many 

 years we have arrived in our laboratory at the following method of exam- 

 ination 2 ): — 



"A piece of filtering paper measuring about 2X2 l /*in., and folded 

 in the shape of a spill, is saturated with the oil, the excess of oil is 

 thrown off and the paper is placed in a small porcelain dish which in 

 its turn reposes in a larger dish, measuring about 8 in. in diameter. 

 The paper is now ignited, and a beaker of a capacity of about 2 litres, 

 moistened inside with distilled water, which has been held in readiness, 

 is quickly placed over the paper. The respective sizes of the inplements 

 used must be such that the rim of the larger dish must stand out well 

 all round from the beaker. After the flame is extinguished the beaker 

 should be left in position for about one minute, after which the products 

 of combustion which have been deposited against the moist sides of the 

 beaker are rinsed out with a little (10 cc.) distilled water, and poured 

 upon a filter. The filtrate, acidulated with a drop of nitric acid, must 

 remain clear when silver nitrate solution is added." 



When this method is followed the test is extraordinarily sensitive, 

 1 drop of monochlorobenzene in 50 g. benzaldehyde (that is to say a 

 very much weaker solution than that used by Herzog), and an even 

 smaller proportion of chlorine being detected without any difficulty, while 

 the same solution when tested by Herzog's method in the presence of 

 a very minute proportion of chlorine in the benzaldehyde was either 

 uncertain or gave a negative result. This is quite natural, because in 

 the last-named case less benzaldehyde is burnt up than with our method. 

 Another advantage of our method is that practically no paper is burnt, 

 the only body which goes into combustion being benzaldehyde. Experi- 

 ments have shown us that many filtering papers contain traces of chlorine, 

 and when therefore the paper is burnt up completely (as required by 

 the German Pharmacopoeia) it is quite possible that a faint chlorine- 

 reaction will make its appearance, which might erroneously be attributed 

 to the faulty condition of the benzaldehyde. We have frequently tried 

 this experiment in our laboratory. We desire to point out further that 

 the beaker does not become heated to any extent if our method is strictly 

 adhered to. 



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



