SHORT AND SIMPLE METHOD FOR DETERMINATION OF VANILLIN. 163 



filter to the fifty cubic centimetres mark in a Nessler glass. 

 Ten drops of a freshly prepared saturated solution of 

 bromine water and ten drops of a ten per cent, solution of 

 ferrous sulphate are added in the order mentioned. The 

 colour is compared with a 0*2 per cent, solution of pure 

 vanillin (tested by titration with decinormal alcoholic 

 potash), either by simple nesslerising or, preferably, with 

 a Duboscq colorimeter. This process is very simple and 

 expeditious, and if carried out carefully, so accurate, that 

 I feel assured it will prove a boon to those whose business 

 it is to standardise vanilla essences. It may be further 

 -simplified by adding the bromine and iron solution to the 

 diluted essence, using one cubic centimetre to fifty cubic 

 centimetres of water, comparing with a similar essence 

 whose vanillin content is known. 



The extraction apparatus shown in Plate VII, is an 

 application of the continuous automatic principle found in 

 the Soxhlet tube, designed for the extraction of liquid, and 

 has been improvised from forms which are usually found in 

 all chemical laboratories. It should be noted that the 

 siphon tube should extend to about half an inch above the 

 liquid to be extracted, and be turned upwards at the end* 

 The size of the separator may be varied to suit the quantity 

 of liquid operated on. Though no originality is claimed for 

 the central idea of this apparatus, I have not seen it else- 

 where in the form described. 



A FLAME TEST FOR CHLORAL HYDRATE. 

 By W. M. DOHERTY, F.I.C., F.C.S. 



[Read before the Royal Society of N. S. Wales, October 1, 1913.] 



I have frequently observed that when chloroform was 

 being evaporated, the flames of the brass Bunsen burners 



