ON THE NATURE OF FERMENTATION 347 
only three days earlier, that the milks of this last experiment were all pure. 
The truth is, all but two have organisms in them ; but I may mention that all 
but four had obviously organisms in them before I went for my trip on the 
Continent three weeks ago. On my return I found that in the course of the 
three weeks that had elapsed, two others had gone ; Hut they already showed 
organisms which, though very pale and insignificant, were quite easily seen 
by a magnifier in such considerable mass that I felt sure they must have already 
been growing for a considerable time ; and, therefore, in all probability those 
that still seemed to the naked eye and to the magnifier free from organisms 
were really so. Accordingly, two days ago I drew out milk from one of those 
that seemed to be still pure, and I had the great satisfaction of finding it not 
only perfectly fluid and tasting perfectly sweet, with a perfectly normal 
reaction, purpling both blue litmus paper and red litmus paper—the normal 
reaction of perfectly fresh milk—but under the microscope I could not dis- 
cover any organism of any kind whatsoever. Therefore, I think we are 
justified in saying that in unboiled milk as in boiled milk, provided, of course, 
the cow be healthy, there does not exist any constituent having the power 
of giving rise to organisms or producing the lactic or any other fermentative 
change. 
This, gentlemen, was the first step of the investigation: to the second 
I must beg your special attention, because I believe you will agree with me 
that it is by far the more important step of the two. 
The object of the second part of the investigation was to find absolute 
evidence, if possible, whether the Bacterium lactis was or was not the cause 
of the lactic fermentation. It occurred to me that, 1f we could estimate with 
some degree of accuracy the number of bacteria present in a given quantity 
of souring milk, and then if we were to dilute the milk with a proportionate 
quantity of boiled water, we might have the diluted milk so arranged that every 
drop with which we should inoculate a series of glasses of boiled milk might 
contain, on the average, one bacterium ; and if we should do so, as it would be 
practically certain that the bacteria would not be distributed with absolute 
uniformity, we should expect that we might have, as the result of these various 
inoculations, some glasses with the Bactertum lactis, and some without it ; and, 
if it should turn out that all those glasses which contained the Bacterium lactis 
underwent lactic fermentation, and, on the other hand, those glasses which 
were free from bacteria had no fermentation, that would prove the point ; as, 
I think, you will agree with me, when we come to discuss the matter at a little 
more length after we have all our facts before us. Well, how were we to deter- 
mine the number of bacteria existing in the liquid? This was done in a simple 
