shall be kept in a well-stoppered bottle, shall be made up freshly each week.
Method.-An appropriate volume of sample containing 0~ooo04-o~ooo1z g. sulphur as carbon disulphide, determined by a preliminary test, shall be pipetted into a 25 ml. graduated cylinder provided with a ground-glass stopper. One millilitre of diethylamine and I ml. of the 0.03% w/v copper acetate solution shall then be added. The volume shall be made up to 25 ml. with carbon disulphide-free benzene and the mixture shaken. After standing 8 min., the mixture shall be transferred to a Nessler cylinder and the colour matched against the colour standards, by viewing the cylinders vertically whilst they are held 3 in. above the surface of an opaque glass sheet reflecting diffused daylight. The final match shall be made 10 min. after the reagents were mixed. The percentage by weight of sulphur as carbon disulphide shall be taken as
C
where C = g. sulphur as carbon disulphide in the colour standard V = volume of sample used. p = the specific gravity of the sample at the temperature The test shall be carried out at a temperature of zoo -f 5". It is hoped that permanent colour standards may eventually become available by co-operation with The Tintometer Ltd. The test can then be modified by matching the colour of the test mixture against appropriate glasses mounted on a disc and viewed in the B.D.H. Lovibond Ncssleriser.
Lbrits of error.-The error in estimating the carbon disulphide will depend upon the quantity of sample used for estimation, but the quantity of carbon disulphide in this volume should be within o-ooooo5 g. of the true value.
Note I .-Carbon disulphide-free benzene shall be prepared from pure ltenzole for nitration* by refluxing it for one hour with one-third of its volume of alcoholic potash. After washing three times with a similar volume of water, the benzene shall be dried with calcium chloride and refractionated through a xz-pear column, the fraction distilling at 80" to 80.5' being collected. Twenty-five millilitres of this benzene, mixed with I ml. of diethylamine and I ml. of a 0.03% w/v solution of copper acetate shall not be darker in colour than a similar quantity of benzene alone, compared in the standard Nessler cylinders.
Note 2.-If it is only desired to see whether a sample contains greater or less than a specified quantity of carbon disulphide, only one colour standard is required.
My thanks are due to the Executive Committee of the National Benzole Association for permission to publish this work.
Research Department, I00 -" X P most nearly matching the sample.
of the test.