294 Daniel D. Jackson 



sists of obtaining the percentage of gas formed by bacterial growth 

 in the closed arm of an inverted tube containing a solution of meat 

 extract, peptone and lactose or dextrose. One-tenth, i,'and lo c.c. of 

 water are added to the sterilized media and allowed to stand 48 hours 

 in an incubator at 37.5° C. At the end of this time the percentage 

 of gas formed and the amount absorbed by caustic is obtained. 



Typical growths of B. coli give from 25 to 70 per cent gas, of which 

 from 25 to 40 per cent is carbonic acid and is absorbed by caustic. 

 This so-called "presumptive test" is used extensively in most water 

 laboratories, and in those of the New York City Department of Water 

 Supply, Gas, and Electricity there are tested annually in this manner 

 about 7,000 samples of water. The difficulty with this test lies in 

 the fact that when badly contaminated water is taken as, for instance, 

 a suspension of feces or a strong sew'age, the test often utterly fails, 

 due to an overgrowth of some germ other than B. coli. 



In such cases B. coli is inhibited in its growth by the products of 

 metabolism of certain other species, usually streptococci. This inhi- 

 bition has been noticed by Prescott and Baker,^ Irons,^° and others, 

 and has been a constant source of annoyance to the author in work 

 on water and sewage. The use of sodium taurocholate as suggested 

 by MacConkey to prevent the growth of these inhibiting bacteria is 

 a step in the right direction, but the salt is very expensive and difficult 

 to obtain, and the amount suggested (0.5 per cent), while effective 

 in soHd media, is, in the opinion of the author, too small for liquid 

 media. 



Inasmuch as the foregoing experiments showed an equal effective- 

 ness of the more abundant sodium glycocholate there appeared to be 

 no reason why a mixture of the two bile salts (Platner's crystallized 

 bile) could not be used. This mixture is easy to obtain and cheap 

 enough to use with liquid media in larger amounts than those em- 

 ployed by MacConkey. It is made as follows: 



Concentrate ox bile to one-fourth its bulk; mix with animal charcoal in a mortar 

 to a thick paste and evaporate to complete dryness over a water bath. 



To the dry charcoal bile mixture add five volumes of absolute alcohol. Shake 

 from time to time, and in about half an hour filter off the alcohol. Concentrate 

 the filtrate by boiling; and when cold add ether in large excess and the crystallized 

 sodium salts of taurocholic and glycocholic acid will be precipitated. Allow the 

 precipitate to stand over night in a tightly covered glass jar and decant off the mix- 

 ture of alcohol and ether. 



