ISOLATION AND PREPARATION 17 



sidue but also on the proportion of water with which it combines. It 

 also depends to a large extent on the character of the solution which 

 the soluble constituents of the seed yield, for some seeds contain 

 substances which produce extremely gummy, viscid solutions which 

 render filtration extremely difficult. No general statement can be 

 made as to the proper proportion of solvent, or the methods to be 

 employed in filtering the extracts, as each seed requires special treat- 

 ment. The filtered extract may be subjected at once to dialysis and 

 the dissolved globulin thus separated ; or the total proteins which it con- 

 tains may be precipitated with ammonium sulphate, redissolved in 

 saline solutions and the globulin precipitated by dialysis ; or it may be 

 first subjected to fractional precipitation from ammonium sulphate 

 solutions of definite concentration and solutions of the resulting fractions 

 separately dialysed. 



Extracts of seeds which yield a large proportion of globulin are 

 usually best subjected to direct dialysis, and the globulin which 

 separates fractionally precipitated from saline solutions or fractionated 

 with ammonium sulphate. Those which yield but little globulin are best 

 saturated at once with ammonium sulphate in order that the precipitate 

 may be redissolved in a small volume of water. The proteins soluble 

 in water may be obtained from the filtrates from the precipitates produced 

 by dialysis according to the general plan above indicated for separating 

 them from aqueous extracts. 



The globulins may also be obtained by sufficiently diluting the 

 sodium chloride extract with pure water and passing carbonic acid 

 through it. The degree to which it should be diluted and the com- 

 pleteness with which globulin can be thus separated from the solution 

 depends on the solubility of the globulin in question and should be 

 determined in ea^h case by special experiments. The separation is 

 rarely as complete by dilution as by direct dialysis of the extract, and 

 the amorphous precipitates and voluminous solutions are not so easily 

 dealt with afterwards. 



The extraction may also be made with warm dilute saline solu- 

 tions and the globulin precipitated by cooling the filtered extract as, 

 in most cases, the globulins are much less soluble in cold solutions 

 than in warm. This method of treatment often results in the produc- 

 tion of crystalline preparations which can be afterwards recrystallised 

 from warm dilute saline solutions and brought to a high state of 

 purity. This method of crystallisation was first extensively employed 

 by Griibler (143), and later by many others, for the production of 

 crystalline preparations from the squash seed. Globulins which sepa- 



