408 R. W. CHALLINOR. 



A series of trials were also made with MacMillari's colour 

 testing apparatus. This apparatus is devised for the deter- 

 mination of carbon in steels (Eggertz test) and consists of 

 two tubes with reservoir tops into which slide two smaller 

 tubes, enclosed in a wooden box and supported over an opal 

 reflector. The standard solution and solution to be examined 

 are placed respectively in the larger tubes, and the smaller 

 tubes are lowered till they rest on the bottom ; when 

 viewed vertically they now appear white. The standard 

 tube is then raised till a suitable tint is obtained and is 

 clamped, the other tube is likewise raised till the tint is 

 the same as standard and also clamped. A vertical scale 

 between the tubes shows the relative depths of the columns 

 of liquid which give the same colour and from this the 

 amount of carbon is calculated. 



Some 22 standard solutions containing different propor- 

 tions of nickel and cobalt in 10 cc. were prepared with the 

 idea that if the solution under examination was first of all 

 matched in colour with one of these standards and then 

 compared with it in the apparatus as above mentioned, the 

 amounts of nickel and cobalt present could be calculated. 

 After comparing about 20 of the standard solutions with 

 each other, under various conditions, the idea was abandoned 

 as the results were far from satisfactory, owing chiefly to 

 the variety of colours produced when nickel and cobalt 

 solutions are mixed in different proportions. 



Other experiments showed that when cobalt and nickel 

 solutions are added to each other a neutral tint is produced, 

 and the change in colour near this tint is found to be 

 sharpest when the total weight of nickel and cobalt in 50 cc. 

 of solution, using these particular tubes, is not more than 

 '05 gram., and that at this strength dilution does not appear 

 to affect the colour of either nickel or cobalt solution 

 separately. 



