136 LABORATORY EXERCISES IN BACTERIOLOGY. 



temperature an active fermentation prevails in the water for a number of days, during 

 which time the water is apt to be turbid and foul and its surface covered with froth ; 

 in the course of three or four days, however, the activity of the process subsides and 

 the water again becomes clear and attractive to the taste, the organic substances 

 having been destroyed. 



Great variation in the number of bacteria found in a water is certain to accompany 

 marked differences of temperature; for which reason it is a fairer statement of the 

 condition of the water to announce a mean result from several examinations made at 

 different seasons, than to give out the result of a single investigation. 



It is essential, too, that the transfer of the specimen to the culture medium should 

 follow its collection as closely as possible; otherwise, owing to the favor of the quiet 

 and warmth of the laboratory in which it is standing, there will be induced a great 

 and rapid multiplication of the germs in the sample, vitiating the results of the analysis. 

 Should it be impossible to at once proceed with the cultivation of the water after it 

 has been collected, as where it must occupy some time in transit to the laboratory, 

 it should be packed in ice to prevent this source of error. 



The method of collection must vary with the relative degree of bacterial con- 

 tamination. A small sample of a water rich in microorganisms will in all probability 

 contain a full representation of the flora of the general supply ; when the water is rela- 

 tively pure, the same amount is very likely to represent it uncertainly both numerically 

 and typically. In the latter instance it will be advisable to practise some measure 

 which will insure the concentration of the flora of a large amount of the water, as 

 upon some form of sterile filter, the washings from which may thus be taken as repre- 

 sentative of the whole amount filtered. 



Water in which currents prevail is likely to have its bacteria and other particles 

 well disseminated through the whole bulk, and a single sample is therefore usually a 

 fair example of the whole. When water is stagnant, however, the upper and lower 

 strata are usually particularly rich in bacteria, the former with those in active growth 

 near a free supply of air, the latter from sedimentation from the higher strata and 

 with bacteria developing in the dead organic deposit and away from a free oxygen- 

 supply. In the latter case, as in cisterns or infrequently used wells, it should be a 

 rule to take samples from different depths, top, middle, and bottom, the report 

 of results being the mean of the series. So, too, when water from a tap is to be ex- 

 amined, because of the tendency for similar sedimentation through the relatively 

 quiet column of water in the ordinary house pipes down to the first free current in 

 a neighboring main, it is the usual practice to let the water flow freely for at least a 

 half hour before the sample is taken. 



It is advantageous, if there be no reason to prevent and no basis for reasonable 

 opinion as to whether there be large numbers or few bacteria in a sample to be ex- 

 amined, that a rough preliminary test be made by inoculating a tube of liquefied 

 gelatine with one cubic centimeter of the natural water, diffusing and plating it. In 

 two or three days one can easily determine the need for concentration of the specimen 

 or for its dilution as there appear few or many colonies. If it be found that few or 

 no growths follow this preliminary test, a filter is arranged, to be attached to the tap 

 after the water has been running from it for a proper time (or to a large sterile funnel 

 into which the water is poured if taken from some source without pipe connection). 

 There is selected a glass or tinned iron cylinder about twenty or thirty centimeters 

 in length and one and a half or tw T o centimeters in the inside diameter. In the interior, 

 as shown in the accompanying diagram (Fig. 41), the folloAving layers are arranged: 



