20 STORRS AGRICULTURAL EXPERIMENT STATION. 



four or five daj'S old shows a differentiation which is clear and 

 sharp, so that it is possible to determine, not only the total 

 number, but approximately the number of each kind of bac- 

 teria. Plates that are older than these are even more useful, 

 because the colonies become more and more distinct in their 

 characteristics the older they grow. 



This necessity of keeping the plates until they are several 

 days old introduces the most serious fault of our method of 

 study. The presence of liquefying bacteria is sometimes quite 

 fatal to the success of an experiment. If a sample of milk 

 chances to contain a considerable number of liquefiers, or even 

 a small number of certain species of rapid liquefiers, the gelatin 

 of the plate may be completely liquefied before the colonies 

 have grown large enough to be differentiated. Under these 

 conditions the experiment becomes a blank, and this difficulty 

 of the liquefying bacteria very seriously interferes with the 

 complete success of the method. 



We have tried to meet the difficulty in several ways. We 

 have tried to use agar instead of gelatin, but this is quite 

 unsatisfactory inasmuch as the colonies are not differentiated 

 on the agar as they are on the gelatin, and it is quite impossible 

 to separate the species from each other on the former as they 

 can be done on the latter. We have made a culture medium 

 of agar and gelatin mixed; but while this makes possible a 

 satisfactory differentiation, it does not so check the liquefaction 

 as to be an improvement over pure gelatin. We have tried to 

 put into each liquefying colony, as soon as it appears, a drop of 

 a weak solution of formaldehyde to stop the growth of the bac- 

 teria, but this is futile, for the enzymes which have been pro- 

 duced continue to spread, and the liquefaction goes on as if the 

 formaldehyde had not been added. Indeed, we have as yet 

 been unable to meet this difficulty and as a result it happens 

 occasionally, as will be seen on the following pages, that an 

 experiment is quite blank because of the rapid growth of lique- 

 fiers, while in other cases the presence of liquefiers increases 

 the number of colonies which are classed as undetermined. 



DIFFERENTIATION OF SPECIES. 



Milk to be studied is diluted by sterile water to an amount 

 which previous experience has shown to be most likely to be 

 useful. In the study of fresh milk as obtained from the milk 



