426 MEMOIES OF THE NATIONAL ACADEMY OF SCIENCES. 



BACTERIOLOGICAL METHODS. 



The -prater for the bacteriological determination was collected in gronud glass stoppered 

 bottles of about 300 cc. capacity. These were previously sterilized by keeping them in the steam 

 sterilizer for an hour or longer. 



For the estimation of the number of l)acteria in 1 cc. Petri's dishes or plates were used and 

 nutrient gelatin (10 per cent) of a neutral or slightly alkaline reaction has been the culture medium 

 usually employed. The plates were prepared from the samples in nearly all cases within about an 

 hour after their arrival at the laboratory. In the majority of instances two or more preparations 

 were made from each sami)k'. The quantity of water taken, often after having been diluted with 

 sterilized distilled water in known proportions, was measured by means of a small sterilized 

 pipette graduated to tenths of a cubic centimeter. The amount of the original water thus used 

 has varied from 0.01 to 0.3 cc. 



When the colonies are counted a hand lens and a glass plate ruled in sqiiare centimeters are 

 used, and with probably two exceptions the numbers given in the table of analyses are based on 

 actual count of the whole plate. The estimation of the number of bacteria to the cubic 

 centimeter in a sample of water is subject to many errors, and the results obtained from two 

 plates made from the same sample of waters at the same time may differ widely from one another. 

 At best the estimations are only apjiroxiniations. There are a number of reasons for this, some 

 of which are as follows: 



(1) The bacteria exist in the water as solid particles in suspension, and it can not be 

 assumed, as in the case of a solution, that there is equal distribution throughout the whole volume 

 of the liquid. It is therefore to be expected that there will be differences in determinations made 

 under the same conditions from a given specimen of water. 



(2) There is an error in measuring out the cpiantity of water which is mixed with the gelatin, 

 most marked when small quantities are measured, such as 0.1 cc. 



(3) Some of the bacteria remain behind with the residue of gelatin left in the tube after the 

 plate has been made, so that not all of the organisms contained in the quantity of water taken 

 are represented by the colonies which later develop. 



(4) Not all of tlie bacteria develop under the conditions ordinarily surrounding them in the 

 medium employed. Some nmy develop slowly, and their colonies may be obscured by more 

 rapidly growing or rapidly liquefying organisms, so that a plate may be destroyed before many 

 colonies have become visible. ^Moreover, it would appear from the work of Reinsch (Centralbl. 

 Bakt. u. Parasit., Bd. X, 1801, p. 115) and of Dahmen (Centralbl. Bakt. u. Parasit., Bd. XII, 1S02, 

 p. .302) that the degree of alkalinity of the culture medium has an effect on the results obtained 

 in these estimations. 



(5) The number of colonies varies with the age of the plate, for some species grow more 

 rapidly than others. 



(6) The results may be vitiated by the development of colonies of bacteria derived from other 

 sources. 



The following table is given as illustrating some of the foregoing statements. It shows some 

 of the differences observed in the number of colonies in plates made with equal quantities of the 

 Game water and at the same time, and also the variation in the counts, made at different times, of 

 ii given plate : 



