372 EXPERIMENT STATION RECORD. 



"(2) The greater part of these proteids consists of a f>;lobuliu, the legumiu of 

 Braconnot, which is readily precipitated by dialyzing its salt solutions. 



"The prevalent idea that legumiu is 8(duble only in acids and alkalies is errone- 

 ous, it having been proved, notably by Kitthausen, to be a true globuliu. The com- 

 position of legumin, as shown by the average of our accordant analyses of 31 

 preparations obtained from the seeds of peas and vetches, is the following: 



Per cent. 



Carbon 52. 15 



Hydrogen 6.96 



Nitrogen 17. 98 



Sulphur 43 



Oxygen 22. 48 



100. 00 



"Legumin is abundantly soluble in solutions containing above 5 per cent of 

 sodium chlorid; in those containing less salt it is not so soluble, the amount held in 

 solution decreasing as the salt content diminishes, so that it is but sparingly solu- 

 ble in solutions containing less than 1 per cent of salt. By dilution with water, 

 strong saline solutions of legumin are abundantly precipitated, 



"By saturation with sodium chlorid or magnesium suljihate, its sodium chlorid 

 s(dutions are not precipitated; by saturation with sodium sulphate at 25° they are 

 not precipitated, but at higher temperatures more or less is thrown down, and by 

 saturation with sodium sulphate at 34", precipitation is very nearlj' complete. 

 With nitric acid and Millon's and Adamkiewic's reagents it gives the usual proteid 

 reactions. 



"With strong solutions of legumin the biuret test gives a violet color at tirst, 

 which on stauding becomes crimson red, similar to the color produced by peptones. 



"The legumin obtained by us from the vetch is not coagulated by heat nor even 

 rendered turl)id by prolonged lioiling of strong solutions. 



"The legumin prepared l)y us from the pea is partly coagulated by heating strong 

 solutions in a boiling water bath, and sets to a firm jelly after thus heating for some 

 time. These ditterences in their behavior on heating, and a greater tendency of the 

 vetch legumin to cohere in semisolid lumj^s when precipitated by dialysis, are the 

 only points of dissimilarity which a rigid comparison of preparations from the 2 

 seeds has revealed. 



"These differences, in our opinion, are due to the substances with which the 

 proteid is associated in the 2 seeds, for saturation of the -pea, extracts with sodium 

 chlorid, before preciiiitating the legumin by dialysis, greatly diminished the amount 

 of coagulum given by the pea legumin. 



"(3) Besides the legumin, the pea and vetch contain another proteid in small 

 amount, either an albumin or a globulin, soluble in extremely dilute salt solutions, 

 and coagulated by heating its solutions to 80^. This substance we have not studied 

 further than to make 2 preparations for analysis from the pea and 1 from the vetch. 

 These were obtained in an insoluble form by coagulating with alcohol, so that the 

 properties and reactions were not determined. The composition of this proteid is 

 shown by the following average of 3 closely agreeing analyses: 



Proteid of pea and vetch. 



Per cent. 



Carbon 53. 48 



Hydrogen 6. 89 



Nitrogen 16. 43 



Sulphur 1. 01 



Oxygen 22.19 



100. 00 



