426 



SCIENCE 



[X. S. V0L.XXVIII. No. 718 



tory. The nmoniit of ammonia yielded by 

 h.ydrolysis can be determined with sxich ae- 

 cnracy that differences of only a very few 

 hnudrodths of a per cent, occur between 

 determinations made on difi'erent prepara- 

 tions of one and the same protein. These 

 fonr determinations are the most reliable 

 that w'e now have for comparing proteins 

 with one another, and make it possible to 

 detect diiferences between them ^Yhich 

 would otherwise escape notice. 



Glutaminic acid can be determined in 

 most cases with a close approximation to 

 its true amount, biit there are some pro- 

 teins, especially those from leguminous 

 seeds, from which it is not easily obtained. 

 Experience has shown that by the ester 

 method alone about seventy-five per cent. 

 as much glutaminic acid is iisually ob- 

 tained as by direct separation as the hydro- 

 chloride. It is possible, therefore, to con- 

 trol to a certain extent the results of direct 

 determinations by comparing them with 

 those obtained on a larger scale by the 

 ester method. 



Although the methods available for thus 

 quantitatively analyzing the products of 

 protein hydrolysis leave much to be de- 

 sired, it must not be forgotten that only 

 recently have we been able to make any 

 comparison whatever of tlie proportion of 

 these products. 



A striking feature of these analyses, to 

 which Professor Chittenden directed atten- 

 tion in his address before this society last 

 January, is that the total quantity of the 

 substances determined falls far short of 

 one hundred j)er cent. The majority of 

 successful analyses foot up between sixty 

 and seventy per cent., and of this a part 

 is made up of water which has been intro- 

 duced by hydrolysis. Calculation shows 

 this amoiint of water to be approximately 

 equivalent to the losses that may be as- 

 siimed to occur through incomplete esteri- 

 fication and separation of the acids, so that 



the sunnnation of well-conducted analyses 

 may be taken as representing somewhere 

 near the total quantity made up by all the 

 different substances determined. 



Nothing is known of the undetermined 

 residue which forms from twenty-five to 

 thirty-five per cent, of the protein. There 

 is no reason to believe that, in the seed pro- 

 teins, undetermined carbohydrate forms 

 anj"- part of this, for those proteins which 

 give no Molisch reaction give no higher 

 sunnnation than do those that do. 



If the amount of nitrogen contained in 

 the quantities of the amino-acids stated in 

 the analyses is subtracted from the total 

 nitrogen of the protein, it is found that this 

 undetermined nitrogen forms about four- 

 teen per cent, of the undetermined part of 

 the protein. This is a higher proportion 

 of nitrogen than is found in any of the 

 monamino-acids that are known to be 

 yielded by proteins, except glycocoll, 

 alanine, serine and tryptophane, even if 

 the proportion of nitrogen is calculated for 

 them as united with one another in poly- 

 peptide union. It is improbable that this 

 undetermined residue is made up of these 

 four amino-acids. and we may expect to 

 find still undiscovered substances among 

 the protein decomposition products. 



COMPARISON OP THE PROTEINS OP DIPPERFJ^fT 

 SEEDS 



The results of this comparative study of 

 the seed proteins shows that no two seeds 

 are alike in respect to their protein con- 

 stituents. Similar proteins are found only 

 in seeds that are botanically closely re- 

 lated. 



The cereals are alike in the proportion 

 and general character of their proteins. 

 The seeds of each of these, with the pi-ob- 

 able exception of those of rice, contain a 

 small amount of proteose, albumin and 

 globulin, and relatively considerable quan- 

 tities of prolamin soluble in alcohol, and 



