Kirchner — Bacteriological Examination of Biver Water. 269 



adjusted around the stopper, the apparatus may be lowered 

 into the water. By pulling the rod the stopper will be raised 

 and the water will enter the bottle. On releasing the hold on 

 the rod, the spring returns the stopper into the neck of the 

 bottle. The stopper of the tube which contains the ther- 

 mometer having also been removed, the apparatus may be 

 withdrawn from the water. 



The apparatus has the advantage that it may be operated 

 with one hand- It is firm and simple of construction, and, 

 although subjected to rough usage, has given entire satisfac- 

 tion. A similar but larger apparatus has been used for 

 taking samples for chemical analysis. 



The bottle for the collection of samples should have a well- 

 fitting stopper, and the top having been wrapped with lead- 

 foil and a cloth, the bottle should be sterilized. After 

 collecting, the lead-foil and cloth should be replaced, and the 

 bottle should be placed in a clean can and packed in ice. 

 Each sample should be properly labeled, and should bear the 

 date, location, temperature of air and of water, name of col- 

 lector, and other special data. 



Transportation of Samples. — The samples having been 

 collected, they should be examined at once, or, if this is not 

 possible, they should be transported to the laboratory with 

 the greatest possible dispatch. It is essential that the samples 

 be kept not only " cold," but their temperature while in 

 transit should be kept at or near the freezing-point by packing 

 them in ice. It is well-known that bacteria in samples of 

 water generally multiply and reach enormous numbers under 

 favorable conditions of growth, but in order to obtain definite 

 information on this point, I made a number of tests which 

 illustrate of how great importance the factors of time and tem- 

 perature are in arriving at the actual conditions of the water. 



Samples arriving at the laboratory were plated immediately 

 for a quantitative determination and were then placed in that 

 compartment of a household refrigerator where the ice is 

 kept. The samples were replated on the next day after an 

 interval of nineteen to twenty-four hours. Sixty-six tests 

 were made in this manner, but for illustration it will be 

 necessary only to give a portion of the results. (Table I.) 



