EM PREPARATIONS AND ‘Drugs. Ren neu.) 68 
‘condenser, with 1 g. pivtassiurh hydroxide free cout antorine. ‘dissolved i in 20 cc. 
fa cohol. The mixture is. then well shaken with 40 cc. of water, separated in a 
x: arating funnel and the remaining benzaldehyde shaken again with 40 cc. of water. 
\ but. 100 cc. of liquid are thus obtained, which are acidulated with 2 cc. of concentrated 
litric acid. We then add an N/10 (in case the chlorine content is extremely feeble 
N ‘/° 00) ‘solution of silver nitrate in excess and 2 cc. of an iron- alum solution’), after 
which we extract with, ether. As- soon as. the ether has separated again, the watery 
layer is withdrawn and the excess of silver nitrate titrated back with a N/10 or a N/100 
olution of ammonium sulphocyanide, as the case may be. 
_ Of strongly chlorinous benzaldehydes only 10 g. need be used, when of course all _ 
n 4 B other quantities indicated must be reduced to half. It is easy to estimate roughly 
> chlorine content by the qualitative test (combustion method)’). 
‘The following values were established OY this: method for mixtures of known: 
| icine content: * 
_Benzaldehyde 0.1.5 p. ce Cl (added in form of benzyl chloride) found 0.096 pic: ch 
” ++ 0.05 p.- C. 5 ( ” ” Yo an : - )) | if ~ 0.045 p-.c. 5, 
ee OM eo eS he net a ae end Cis 2 OTB py ex, 
Ca mae + 0.01 Boe tiie ea We a et yet 0000: sp. C25" 
>This shows that the results obtained are quite satisfactory, but the process is after 
I rather complicated, which has all the more weight, as it does not permit of determining 
> chlorine in the. ‘nucleus, as stated before, upon which everything here depends. 
I setinore: there is the other drawback, that it may easily give rise to differences 
oretaste thereof, when we procured for ourselves three samples with guaranteed 
sontent of chlorine. Only in one case the quantity of chlorine determined by above | 
aethod tallied with that indicated, whereas in the two other cases there were rather 
co: 01 nsiderable differences: — © Peake SN aon 
- Benzaldehyde I. guaranteed: to contain 0.2, found 041 per cent.. 
set metas AL ¥: Re eye oer Oy os he OLOBT ae ia 6 
ne Ill. “oh rpg aye O02 oi FP OOBE ys k 4 
Iti is almost the same with Vaubel’s test (heating the benzaldehyde with concentrated 
Bitphicic acid. Comp. Report April/October 1918, 61), which M. S. Salamon recommends 
as eee Only the chlorine of the side chain can be determined with it. 
As matters now stand, we are of opinion ‘that it is worth considering, whether 
Be sca not be better to adhere to the former stand-point, according to which a 
3 aldehyde is either free from chlorine, or not. If the article free from chlorine 
Ss extremely scarce at present, it is a sad sign of the times we are living in, but we 
pit t hope for better ones. The essential question is after all, whether chlorine 
S present, or not, and. whether there is a little more or less, comes second. As long 
as there is no really reliable method, the quantitative test ought to be restricted to 
' 5 e it is _absolutely necessary, whereas. in other cases one ‘should take as a criterion 
e result of the qualitative chlorine test according to the combustion method. If the’ 
escription is strictly adhered to, this test permits of forming a bie whether the 
= The iron-alum solution must not be aided after the aon. as there: would again be a Spcaiaeltan 
sa solution of iron-alum in water, saturated in the cold, to which. concentrated mitric acid has been added 
il 1 it is colourless. The nitric acid must be free from nitrogen: oxides, as these would disturb when titrating 
% with a solution of ammonium sulphocyanide. They are eliminated by boiling. —.*) Report October.191}, 
i Gildemeister and Hoffmann, The Volatile Qils, 24 ed., vol. 1, p. 609. tag ay hey ye 
5 
7 
bet een seller and buyer, when the chlorine content is guaranteed. We had a little 
