ABSTEACTS 113 



under dusty conditions occasioned by feeding hay or corn stalks gave 

 noticeably higher figures than any of those recorded above. The 

 results of these four analyses were 1100, 2400, 3957 and 16,070 per 

 liter respectively. 



From the foregoing results, it is evident that the air of dairy stables 

 contains many more bacteria than have been found by Winslow and 

 Browne {Monthly Weather Bureau, 42: 452^53, 1914) in country air, 

 city street air, offices, factories and schools. This is not surprising as 

 relatively dusty operations such as feeding dry hay, grain and the 

 like must be carried out several times daily in every cow stable. In 

 spite of this fact, it must not be concluded that air plays a great part, 

 numerically, in the contamination of milk by bacteria. The studies 

 made in order to discover the relative importance of this factor in 

 milk contamination have shown that the air is a relatively unimportant 

 source of bacteria in milk. The detailed results of the latter investi- 

 gations are published in Bulletin 409 of the N. Y. Agr. Exp. Sta. which 

 has just been issued. 



Validity of Presumptive Tests. W. F. Monfort. 



Each presumptive coli test proposed from time to time has been 

 first applied locally. Its extension to other regions and other classes 

 of waters has developed certain limitations. None has proved of 

 universal application. It is therefore of first importance that the saving 

 clause of Standard Methods, 1912, page 96, be given due consideration 

 in evaluating any abridged test for the colon group before its adoption 

 with respect to waters of a class or region new to the investigator. 

 This discrimination is not always practiced. 



There follow some results of such an evaluation with respect to a 

 surface water (Missouri River) from which a turbidity of 1000 to 12,000 

 parts per million has been removed; the effluent is treated with bleach. 

 Observations covering more than two years show this supply to con- 

 tain usually not more than two organisms of the colon group per 100 

 cubic centimeters. 



Neutral red bile-salt lactose broth gives positive results in all dilutions. 



Aesculin bile-salt broth, giving negative results with B. cloacae, 

 yields a brown coloration with an organism of frequent occurrence 

 belonging to the class B. fluorescens. 



Lactose-bile gives an error of over 73 per cent as compared with 

 confirmation tests of lactose-fermenting, acid-forming, aerobic bacteria. 



In lactose broth 80 per cent of gas formers fail of confirmation. 



For a water of this class apparently nothing thus far proposed short 

 of actual discriminatory tests, at least so far as outhned in the lately 

 adopted "standard" for waters used on common carriers in interstate 

 commerce, can be considered valid. 



