368 ON THE LACTIC FERMENTATION 
be mere accidental concomitants of the fermentative changes ; and when I had 
finished my story he said, ‘ Well, I am convinced.’ I thought to myself, ‘ If 
I have convinced this eminent teacher by these facts, they are worth proving 
more rigorously ; I shall have a little time between giving up my surgeoncy 
in Edinburgh and going to London, and I will devote it to repeating the experi- 
ments ; but this time I will make continuous observations under the microscope, 
and trace, if I can, the actual process of transformation of the organism from 
one form to another.’ Accordingly, I got some souring milk from the same 
dairy as before, and proceeded, in the first place, to inoculate from it a glass 
of uncontaminated unboiled urine, and also a glass of pure Pasteur’s solution. 
The result in the case of the urine was that, instead of getting a large motionless 
spirillum as in the corresponding experiment of four years previously, I found 
an active double bacterium of moderate dimensions. In the Pasteur’s solution, 
on the other hand, instead of the minute very actively moving bacterium of my 
former experience, I obtained only motionless bacteria of various dimensions. 
Here then were the facts all wrong. What was the explanation? I had 
obviously got some accidental contamination of the Bacteriwm lactis with other 
forms, although in doing the inoculations I had only introduced a very minute 
portion with the point of a heated needle. So I determined to try if possible 
to get rid of concomitant bacteria of other kinds, and the way which occurred 
to me as a possible mode of doing this was to dilute the souring milk with so 
large a quantity of boiled and therefore pure water as to have on the average, 
so far as it could be estimated, only one bacterium of any kind to every one 
of the drops with which a set of glasses of boiled milk should be inoculated. 
If this could really be done, as the Bacterium lactis was certainly in much larger 
numbers than any other kind, I might hope that some at least of the drops of 
inoculation might contain it isolated from other species, and that thus I might 
have the Bactertwm lactis develop pure and unmixed in the inoculated milk. 
Accordingly having obtained some souring milk from the dairy, I inoculated 
a glass of boiled milk from it by dipping the point of a heated needle successively 
in the two liquids, and when the odour of souring milk was perceptible in the 
air under the glass shade, I found bacteria present on microscopic examination, 
and endeavoured to estimate their numbers in proportion to the liquid. 
This was done in the following manner. By means of the syringe already 
described (p. 364) one or more hundredths of a minim could be measured with 
precise accuracy ; and I found that 1-50th minim exactly occupied a circular 
plate of thin covering glass, half an inch in diameter, so that when such a drop 
was placed on a glass slide, and a covering glass of the size mentioned and quite 
flat was put down upon it, all air was expelled from under the latter, and the 
