to the remainder and set in the Arnold sterilizer for 15 minutes. Such 

 treatment usually results in an increased acidity, the PH changing to a 

 value between 7.7 and 7.5. This preparation had a final PH = 7.7. The 

 medium was filtered and sterilized in the "Arnold" for 30 minutes on three 

 successive days. 



(b) Carbohydrate broth: The chief point to guard against in this connec- 

 tion is that a sugar in slightly alkaline solution decomposes readily. This 

 occurs to some extent even at room temperature (23 to 25 C.) but very 

 marked decomposition sets in at the temperature of the Arnold sterilizer. 

 It is best, therefore, to sterilize a concentrated (20 per cent) aqueous so- 

 lution of the carbohydrate by boiling (10 to 15 minutes being sufficient) 

 (20). This concentrated solution is then mixed with sterile sugar-free 

 broth in amounts to make the desired dilution. The sterility of the mix- 

 tures was tested by incubating them for 18 hours at 37 C. 



(c) Meat infusion agar: The basis for this medium was made by in- 

 fusing 75 grams Bacto-veal in 500 cc. tap water in the "Arnold" for two 

 hours. The steps in the procedure for obtaining the finished bouillon 

 product parallelled those for sugar-free broth with the exception that 

 there was no inoculation with B. coli. To this substrate was added an 

 equal quantity of a 4 per cent solution of agar-agar, giving a final con- 

 centration equal to 2 per cent. 



(d) Meat infusion gelatin: Hiss and Zinsser recipe (22). 



(e) Milk: Hiss and Zinsser recipe (23). 



(f) Hiss inulin-serum-ivater : Hiss and Zinsser recipe (24) . 



(g) Bile: Ox-bile obtained from the slaughter house was sterilized for 

 20 minutes on three successive days. 



3. PREPARATION OF THE REAGENTS USED IN THE DETERMINATION OF 

 HYDROGEN-ION CONCENTRATIONS: In determining hydrogen-ion concen- 

 trations the colorimetric method was chosen. The principles involved 

 need not be reiterated here; they have been thoroughly described by Clark 

 and Lubs (25). The chief sources of error, color of medium and tubidity 

 of culture were obviated by diluting the medium (one part medium to two 

 parts distilled water) and then compensating for the color by Walpole's 

 comparator method of superimposing the color of the medium upon that 

 of the indicator (26, 27). 



(a) Standard solutions used: i /I5 M disodium hydrogen phosphate 



I /i 5 M potassium dihydrogen phosphate 

 I /5 M sodium acetate 

 i /5 M acetic acid 



The salts were recrystallized repeatedly until they satisfied Sorensen's 

 purity standards (28). The acetic acid was obtained by redistilling C.P. 

 glacial acetic acid. The phosphate mixtures were prepared by Sorensen's 

 technique and the acetic acid-sodium acetate mixtures according to Wal- 

 pole's directions (26). The accuracy of these standard mixtures was 

 verified by the hydrogen electrode.* The actual values obtained prac- 

 tically duplicate those given by these authors. A study of the following 

 table will make this evident. 



*We are indebted to Dr. Hastings, of the Rockefeller Institute, New York City, for 

 aid in this connection. 



