12 CASSAVA. 



METHODS OF ANALYSIS. 



The method of analysis employed in these investigations was dis- 

 tinctly different from that qf previous analysts working along these 

 lines, though it was based upon the same general principle, namely, 

 that hydrocyanic acid can be expelled by heat and made to combine 

 with an alkali, when it forms a definite salt. In detail the method 

 used for the estimation of the hydrocyanic acid was as follows: 



Pulp the fresh tuber, a machine designed for grating horse-radish 

 being used. Place 540 grams of this pulp in a retort without the 

 addition of water and heat gently on an asbestos board over an open 

 flame, continuing the digestion long enough to insure the complete 

 expulsion of the acid about two hours is usually necessary. (The 

 retort used was a pear-shaped iron vessel covered with porcelain 

 enamel and was about three-fourths full when containing the charge 

 specified. This vessel was closed by a No. 10 rubber stopper carding 

 a quarter-inch glass tube which dipped into a solution containing suffi- 

 cient potash to take up all the hydrocyanic acid expelled.) 



The general method at this point is to titrate with a silver salt. 

 There are, however, two sources of error involved in this procedure, 

 one arising from the fact that a minimum excess of potassium in the 

 resultant distillate can not be insured, as the amount of acid driven 

 over is unknown and the accuracy of the titration varies with the 

 excess of potassium hydrate present. A further error and one of 

 much more importance is due to the fact that organic matter is driven 

 over and so discolors the solution that the end reaction is obscured. 

 For these reasons a modification of the method was introduced at this 

 point. 



Place the distillate in a flask, treat with an excess of sulphuric acid, 

 and immediately stopper with a connecting glass tube which dips into 

 water containing about 5 cc of tenth-normal potassium hydrate. 

 Heat the flask gently to drive over the liberated hydrocyanic acid 

 and titrate with tenth-normal silver nitrate as soon as the distillation 

 begins. In solutions of this strength approximately 2 cc of potassium 

 hydrate unite with the hydrocyanic acid and form potassium cyan id 

 with the equivalent of 1 cc of the silver nitrate solution. With this 

 procedure it is easy to provide for a minimum excess of potassium 

 hydrate, as this can be added when the silver nitrate approaches equi- 

 librium. Each cubic centimeter of the silver nitrate used is the 

 equivalent of 1 part per hundred thousand of hydrocyanic acid on 

 the basis of 540 grams of substance used. Duplicates obtained on 

 samples taken from the same lot of well-mixed pulp were almost 

 identical. 



In the use of this method it was observed that the greater part of 

 the hydrocyanic acid is expelled in the early stages of the operation 



