[Vol. 9 



378 ANNALS OF THE MISSOURI BOTANICAL GARDEN 



Beef extract is the remaining nutrient to be considered. Me- 

 dia containing but 1 per cent dextrose and 0.1 per cent tyrosine 

 in agar were tried, and no precipitate was obtained either in 

 acid or in alkaline media. Growth in these was not good, but 

 this is not the factor in preventing precipitate production since 

 the precipitate was produced in 1 per cent dextrose plus beef ex- 

 tract even though growth was poor. It may then be concluded 

 that beef extract is the probable source of the white precipitate, 

 and while no similar study has been made on whey agar it is 



quite probable that whey is the source of the precipitate in this 

 latter medium. 



CHEMICAL NATURE OF THE PRECIPITATE 



The precipitate was soluble in all of the acids tried, including 

 the following: glacial acetic, chromic, citric, hydrochloric, nitric, 

 sulphuric, lactic, picric, and warm boric; it was insoluble in warm 

 or cold water or organic solvents, including methyl, ethyl, and 

 butyl alcohol, glycerin, acetone, ether, petroleum ether, chloro- 

 form, oil of turpentine, benzole, xylol, tuluol, benzine, carbon bi- 

 sulphide, carbon tetrachloride, potassium permanganate, hydro- 

 gen peroxide, and ammonium peroxide. This indicates that the 

 precipitate is an inorganic salt. Tests were made for carbonates 

 and phosphates with the result that posphates were clearly de- 

 tected. Since the precipitate, while conspicuous, may be had in 

 comparatively very small amounts it is rather difficult to test. 

 The procedure was as follows: With a platinum loop the bac- 

 terial smear was carefully removed (all tests were made on agar 

 slants) ; the tube was washed several times with distilled water 

 and care was taken to see that all bacterial growth, where this 

 was a possible factor, as well as particles of medium, were re- 

 moved. One per cent solution of nitric acid was then used to 

 dissolve the precipitate, the fluid decanted, filtered through fine 

 filter-paper, and tested with ammonium molybdate reagent (see 

 Tread well's 'Qualitative Analysis'). Tubes containing the same 

 media but without showing any precipitate, the ones reacting al- 

 kaline, were treated in exactly similar manner and run at the 

 same time. Many such tests show that the precipitate, if not 

 wholly, is in large part phosphate, as indicated by the yellow 

 color, and precipitate developed in solutions coming from tubes 

 which originally contained the white discoloration. 



It is very difficult to obtain a test of any nutrient agar in 

 which the yellow precipitate produced by ammonium molybdate 



