27G Trans. Acad. Sci. of St. Louis. 



It must be clear from the foregoing illustration that unless 

 the samples are collected and transported with the greatest 

 care, the results obtained may be far from the truth. It is 

 usually not possible nor convenient to test samples at the 

 place of collection. I have on a few occasions plated samples 

 on the field and the same samples again at the labora- 

 tory under favorable conditions of transportation, and 

 the results on comparison have always been within the 

 limits of error or variation. Except under favorable 

 conditions, it is better, in routine work, to have all 

 samples sent to the laboratory when it is possible to have 

 them properly shipped in ice. While the experiments on the 

 effects of transportation were carried out independently, the 

 results and conclusions are, for the most part, in accord with 

 those obtained by other observers (Jordan and Irons, 1899; 

 Whipple, 1901). Samples of water that are chilled will often 

 show a decided decrease in numbers, while the same quality 

 of water collected during cold weather and subjected to 

 similar treatment will often show an increase in the number 

 of bacteria. It is also true that waters from different 

 sources do not behave in the same way regarding conditions 

 of temperature. 



Reception of Sample. — The sample upon its reception 

 should receive immediate attention and should not be handled 

 except by the person who is to make the examination. Careful 

 notes should also be made as to the condition in which the 

 sample was received. 



Examination of Sample. — Aside from observations on the 

 physical properties of the water, — odor, color, turbidity, 

 sediment, temperature, etc., the bacteriological examination of 

 water divides itself into: (a) quantitative determinations, (b) 

 qualitative determinations, and (c) special examination. For 

 our purpose, it will be best to take up each of these divisions 

 separately. 



Quantitative Determinations. — By a quantitative determi- 

 nation we mean the estimation of the number of bacteria in 

 a given quantity of water, usually expressed by the number 

 of bacteria per cubic centimeter. At the time that the in- 



