AGRICULTURAL CHEMISTRY AGROTECHNY. 713 



the exact amount of sulphuric acid needed to remove the excess of baryta was 

 added and the barium sulphate extracted with boiling water until no coloration 

 was given by ninhydrin [E. S. R., 26, p. 804]. The filtrate was then concen- 

 trated at 40° C. under reduced pressure until the mother liquor gave no reaction 

 with Millon's reagent. The crude tyrosin which separated was purified by boiling 

 with animal charcoal and fractional crystallization from water until a sample on 

 analysis yielded figures agreeing with pure tyrosin." 



Fonnol-titrametric investigations with proteins, II, F. Obeemayer and 

 R. WiLLHEiM (Biochem. Ztschr., 50 (1913), No. 5-6, pp. 369-385).— Continuing 

 work previously noted (E. S. R., 27, p. 501), the authors emphasize the fact 

 that with the formaldehyde titration method the terminal amino groups of pro- 

 teins can be determined. The quotient obtained by dividing the total nitrogen 

 by the number of such axnino groups, and which shows how much nitrogen be- 

 longs to each amino group, is termed the amino index. 



The index of euglobulin (average 21.5) is much greater than albumin (aver- 

 age 12). In the Mammalia this also applies to pseudoglobulin, whei'eas in the 

 hen the index of pseudoglobulins (average about 15) is not so sharply differen- 

 tiated from albumin. 



With the procedure it can be shown that the large protein fractions do not 

 consist of unit substances; thus with pseudoglobulin, by .salting out with 44 

 per cent of ammonium sulphate and albumin with a 66 per cent ammonium 

 sulphate solution, it is possible to obtain two fractious which differ very 

 much in their amino indexes. 



This work, according to the authors, for the first time shows that a con- 

 stitutional difference exists between two different serum proteins. In the case 

 of the hen the amino indexes of the fractions obtained with a 25 per cent 

 ammonium sulphate solution and those given by a 30 per cent ammonium 

 sulphate solution (average 32.5 and 28.5, respectively) were much higher than 

 those given by the serum proteins of the bo^'ine (average 19 and 21.5, re- 

 spectively). Those of the horse were found to behave in exactly the same 

 manner as in the bovine, and those in the goose like those in the hen. 



Detection of saponin, L. Rosenthalee and H. Schellhaas {Ztschr. Unter- 

 such. Nahr. u. Oenussmtl., 25 (1913), No. 3, pp. 154-158). — Instead of determin- 

 ing the presence of saponin by noting the hemolytic power, the authors recom- 

 mend hydrolyzing this glucosid and testing for the cleavage products. This 

 allows of the detection of nonhemolytic regenerated saponin. The method is as 

 follows : 



To the solution under examination add 2.5 cc. of hydrochloric acid, filter if a 

 precipitate ensues, and evaporate the filtrate on a water bath until the solution 

 ceases to foam. This indicates that the hydrolysis is complete. Then cool some- 

 what and shake the still warm solution with acetic ether, taking .50 cc. of 

 acetic ether for each 100 cc. of aqueous extract. Wash the clear acetic acid 

 solution with water until no reaction is obtained with silver nitrate, and evapo- 

 rate to dryness. If the extract is highly colored, it can be decolorized before 

 evaporation with animal charcoal. 



The extract is used for two tests: (1) For prosapogenin with sulphuric acid. 

 This gives an orange red coloration which slowly goes over to a cherry red and 

 finally becomes violet. When small amounts are present this change in colors 

 may take several hours. (2) To see if foaming takes place a little of the ex- 

 tract is dissolved in a small amount of sodium carbonate solution. 



Beer can be tested by a modified form of the above procedure. 



The determination of the acetyl number [of oils, fats, etc.], E. B. Holland 

 (Massachusetts Sta. Bui. 151 (1914), pp. 69-78; Jour. Indus, and Engin. Cheni., 

 6 (1914), No. 6, pp. 482-486). — After reviewing existing methods for determining 



