« 



1889.] Albumose from the Seeds of Abrus precatorius. 101 



globulin is a powerful toxic agent, it was desirable to ascertain 

 whether the albumose possessed the same power and produced the 

 same symptoms. 



It is very difficult to obtain the albumose in a pure state separate 

 from the globulin. Boiling the solution, of course, readily pre- 

 cipitates the globulin, leaving the albumose in solution, but as heat 

 destroys the activity of abrus^ poison, it cannot be employed in 

 separating the two proteids. Both proteids are also thrown down by 

 saturating their solution with neutral ammonium sulphate. The 

 precipitate thus formed can be redissolved and the solution dialysed, 

 thus removing most of the salt and precipitating the globulin. But 

 I found many objections to this method. The dialysis has to be 

 prolonged over a week, and there is thus great liability to decomposi- 

 tion. Moreover, it is practically impossible to precipitate all the 

 globulin by dialysis, and the ammonium sulphate, traces of which 

 still remain, being itself poisonous, would be likely to vitiate the 

 result in testing the toxic, action of the proteid. I therefore 

 abandoned this method and tried the following, which answered per- 

 fectly. A concentrated watery extract of the seed was made and 

 filtered direct into an excess of absolute alcohol. The copious pre- 

 cipitate which fell consisted of globulin and albumose. After a 

 few days, the proteids were removed by nitration, washed with 

 alcohol, redissolved in water, and reprecipitated by absolute alcohol. 

 They were allowed to remain under absolute alcohol for several 

 months in order to coagulate the globulin, and were then filtered off, 

 redissolved, and reprecipitated by alcohol, and allowed to remain 

 under alcohol for a few months longer. Altogether some of the 

 proteids were allowed to remain under alcohol for eight months, or 

 longer. At the end of this time they were removed, washed with 

 alcohol, and dried over sulphuric acid. The residue was ground into 

 a yellowish -brown powder, and consisted of coagulated globulin and 

 of unaltered albumose. 



For the purpose of inoculation this powder was mixed in distilled 

 water, which had been well boiled to sterilise it and then cooled. 

 The mixture was filtered and the filtrate was clear. It gave the 

 following reactions : — 



1. Neutral to test-paper. 



2. No precipitate on boiling. 



3. Acetic acid gave a precipitate, which mostly redissolved on 

 boiling, coming down again on cooling, and so on. After boiling and 

 cooling, the precipitate was readily soluble in dilute potash, showing 

 that the proteid was not coagulated. 



4. Nitric acid caused a precipitate, mostly soluble on heating, coming 

 down again on cooling, &c. This precipitate, like the acetic acid one, 

 is also readily soluble, after being heated, in dilute potash. 



