The Antitoxins 159 



days, but that the animal receiving that dose, though very ill, lives 

 longer. The test dose may then be assumed to be 0.1, or it may 

 be calculated more closely if desired. 



To test the serum itself, guinea-pigs weighing exactly 250 grams 

 are now all given toxic bouillon 0.1 c.c. plus varying quantities of the 

 serum ^, ^fa, ^fa, etc. All live except those receiving less than 

 T 7 , which die about or on the fourth day. The serum can then be 

 assumed to have 400 units per cubic centimeter unless it be desired 

 to test more closely. 



Standard test serums for making tests of antitoxic serums 

 by the Ehrlich method were first shipped at small expense 

 from the Kaiserliches Institut fur Serum-Therapie at Hochst- 

 on-the-Main. At present the Hygienic Laboratory of the 

 United States Public Health and Marine Hospital Service 

 has legal control of the manufacture of therapeutic serums 

 and kindred products in the United States, issuing licenses 

 to those engaged in legitimate manufacture, and furnishing 

 a standrad test serum, similar to that of Ehrlich, to those 

 entitled to receive it. 



A full description of " The Immunity Unit for Standard- 

 izing Diphtheria Antitoxin," by M. J. Rosenau, Director of 

 the Hygienic Laboratory, can be found in Bulletin No. 21 

 of the U. S. Public Health and Marine Hospital Service, 

 Washington, 1905. 



As the quantity to be injected at each dose diminishes 

 according to the number of units per cubic centimeter the 

 serum contains, it is of the highest importance that therapeu- 

 tic serums be as strong as possible. Various methods of 

 concentration have been suggested. Bujwid * and H. C. 

 Ernst f found that when an antitoxic serum is frozen and 

 then thawed, it separates into two layers, the upper stratum 

 watery, the lower yellowish, the antitoxic value of the 

 yellowish layer being about three times that of the original 

 serum, the upper layer consisting chiefly of water. 



The most satisfactory method of securing a useful concen- 

 tration is by the employment of the globulin precipitation 

 as recommended by Gibson, J which is briefly as follows: 

 The diluted citrated plasma is precipitated with an equal 

 volume of saturated ammonium sulphate solution and the 

 antitoxic proteins separated by extracting the precipitate 



*"Centralbl. f. Bakt. u. Parasitenk.," Sept. 1897, Bd. xxn, Nos. 

 10 and 1 1, p. 287. 



f "Jour. Boston Soc. of Med. Sci.," May, 1898, vol. n, No. 8, p. 

 i37- 



t "Jour. Biol. Chem.," i, p. 161; m, p. 253. 



