302 



again possibilities of castor oil seeds being" mixed with other 

 seeds. A further risk is run at the pressing- factory where the 

 machines may be badly cleaned after pressing the castor oil seeds, 

 so that these become mixed with the next kind of seeds pressed. 

 Again, for a soap-making process in Germany the use of castor 

 oil seeds is necessary, and there is the chance of their getting 

 into animal foods owing to the amount of transport of these 

 seeds that has to be carried on. Lastly, large quantities of the 

 shells are sold at low prices to manufacturers of compound feed- 

 ing-cakes who grind and use these shells in the cakes. As no 

 method is known of completely freeing the shell from the kernel, 

 it follows that these cakes must, as a rule, be poisonous, and on 

 an average Robert estimates that at least 1 per cent of kernel 

 matter will be present with the shell, an amount which is more 

 than sufficient to cause fatal poisoning of cows when, it is re- 

 membered that cakes are fed at the rate of from 2y^ lbs. up to 

 8-)4 lbs. per head per day. Farmers should refuse all such cakes, 

 and merchants who resort to such practice are as guilty as if they 

 included arsenic in their cakes. 



"The poison, ricin, is an albumin and has the characteristic 

 (1) of an albumxin, (2) of a ferment of enzyme, (3) of a toxin, 

 (4) of an agglutinin. 



"From the albumin nature of the poison it results that the mix- 

 ture of ricin with human or animal foods cannot be detected by 

 purely chemical methods, even when one hundred times the fatal 

 dose is contained in the foods ; but the possibility of extracting 

 the poison from foods by water or other method rests on the 

 albumin nature of the poison. 



"The enzyme characteristics of the poison are useless for pur- 

 poses of detection, since feeding cakes are always found to con- 

 tain enzymes similar in effect to ricin. 



"As regards its toxic effects immunity is reached by small and 

 gradually increasing doses ; and in the blood serum of amnuinized 

 animals 'antiricin.' which has the effect of an antitoxin, is formed. 

 This serum has been found extremely effective in the detection 

 of extremely small quantities of ricin, but there is the drawback 

 with this method that a different scrum is j:)roduced in the case of 

 some varieties. 



"The method of detection by injection into guinea pigs and 

 observing wdiether symptoms of super-sensitiveness arc produced 

 is not recommended by Kobert. 



"He lays stress, however, on the efficacy of a third method 

 which rests on the agglutinin rliaracteristics of the poison, i. e.. 

 even if diluted to one-millionth ])art of the original strength it 

 coagulates the blood cor])usclcs of guinea pigs, and a substance 

 like sealing wax is obtained on filtering. This method holds 

 good for all varieties of ricinus and is even more sensitive than 

 the serum test. Even here it must be reincmbercd that 'jihasins' 

 give a similar reaction. Kicin, however, u ill stand a temperature 



