576 



Dr. A. D. Waller. New Method for the [June 17, 



solution of potassium bichromate. It is a very " fast " colour — not being 

 appreciably affected by direct sunlight or by boiling. 



Thus a tint of T 1 in 1 c.c. liquid represents O'OOOOOl gramme HON. 

 T 1 in 15 c.c. „ 0-000015 



T5 in 20 c.c. „ 0-000100 



T 10 in 100 CO. „ 0-001000 



For actual comparison it is preferable to dilute a given liquid, if very 

 highly coloured, so as to bring it below the intensity T 10. For delicate 

 estimations, as e.g. of HON in the blood of an animal poisoned by inhalation 

 of HON, it is possible to distinguish with certainty tints below the intensity 

 T 1 (vide infra, Experiments 11 and 12). 



The rate of development of colour in the reaction between picric and 

 hydrocyanic acid is illustrated by the following figure. At 20° the colour of 









\ \^ 







- \% \C>o 







■ y-** \ 









T5 







T7-5 



I I I I I I I ■ ■ ■ I I I I I I I I I I I I — I — I — I 



0 I i 3 4 



Hours 



Fig. 1. — Curves showing the rate of development of colour, i.e. of the formation of 

 sodium isopurpurate, in a mixture of equal volumes of picrate fluid and of 0-002 per 

 cent. HCN, at 20° and at 40°. 



a mixture of equal volumes picrate and HCN (0-002 per cent.) reached one- 

 half its maximal value in 40 minutes, three-quarters its maximal value in 

 80 minutes. At 40° these times were shortened to 10 and 20 minutes 

 respectively. From which, admitting that the time of intensification to 

 within 1/n of maximum depth of tint varies as the logarithm of n, we may 

 calculate times of intensification to within 1/100 of its full depth at different 

 temperatures. Thus the tint of a mixture should reach to within 1/100 of 

 fulness, i.e. to 99/100, in 4 hours 46 minutes at 20° and in 1 hour 6 minutes 

 at 40°. Distillates should therefore be left in the incubator for at least an 

 hour before their tints are estimated. 



This method, by which it is easy to estimate minute quantities of HCN, 

 counting by thousandths of a milligramme, with an error by manipulation and 

 reading that does not, at present, amount to 10 per cent, of the reading, and 

 that, no doubt, will be reduced by further experience, is very generally 



