[HARRISON A vandkrlfck] .ESCULIN bile SAF.T MEDIA 



161 



media. This gives satisfactory results, because the total amount of 

 aesculin is available, and the depth of the liquid intensifies the colour. 

 Through the latter fact the expense of sesculin tests with large quan- 

 tities of water (1 liter) is not much more than the expense of small 

 tests. 



Several tests were made to see if it were possible to detect 7>. coll 

 in river water and isolate it when only one bacillus was present in a 

 liter of Water. 



This test was conducted as follows: River w^ater was sterilised 

 in large Erlenmeyer flasks, and a small quantity of fresh river water 

 added, w^hich was analysed quantitatively at the same time. To 

 the flasks was added a certain amount of the following medium: 



1000 c.c. distilled water, 1 gr. sesculin, 2 gr. iron citrate, 20 gr. 

 peptone, acidity -f" 1.5. 



Test I. — analyses of river water, No. of colon in 5 cc. respectively, 

 7, 8, 6, 9, 7, average 1^ per cc. 



Black col . 



on sub.s. 



plate.s 



Flask VI. was too diluted to show a black colour; in tlii.s case 

 we had only in one bottle 10 m. gr. aesculin in 1000 c.c. water or 

 0.001%, but flask V. with 0.003% sescuUn was quite effective. 



These results show the reliability of this test, and enable U3 to 

 detect slight infection with B. co/i in potable waters. 



Prejxiration of ^Kscidni. 



Chemically pure sesculin is expensive, but for bacteriological 

 purposes a commercial product suffices, which mayj^be easily prepared. 

 This glucoside is obtained from several plants, but can be most ad van- 



