GERM THEORY OF FERMENTATIVE CHANGES 313 
probably be stated equally truly by saying that, of all the many particles of 
dust that fell in during that period, only three at most contained a germ capable 
of growing in urine. 
Hence while it is most true that scrupulous care must be taken in these 
experiments, and that forgetfulness or slovenliness in their execution would be 
absolutely fatal to success, yet it is equally true that by the very simple means 
which I am now describing the observations may be made with a facility and 
precision that leave nothing to be desired. 
The glass plate and shade are simply washed and dried with a towel, but 
the liqueur-glass and its cap must be purified by heat like the flask. This is 
very simply done by bringing both to a high temperature over two spirit-lamps 
or Bunsen’s burners, the liqueur-glass being held in the hand by its foot and 
the cap in a pair of forceps; and the cap having then been placed on the glass, 
a substantial piece of cotton-wool with a bit of muslin beneath it (neither car- 
bolized) is placed on the cap and tied firmly with fine iron wire round the glass 
beneath. The heat of the glasses ensures the destruction of organisms in the 
part of the cotton, which acts as a perfect filter during cooling; and though 
the muslin may be browned by the high temperature, no empyreumatic odour 
is occasioned in the glass nor any deposit on its sides! The glasses having 
cooled, the wire ligature is cut and the cotton carefully removed, the muslin 
beneath serving the purpose of clearing off all portions of cotton at once, and 
the glass and cap are immediately placed on the glass plate beneath the shade. 
A series of experimental glasses, say a dozen, having been thus prepared, 
it remains to charge them by decanting from the flask, and this is a matter 
which, at the risk of appearing tedious, I am compelled to describe in minute 
detail. 
The process is effected by means of a syphon of glass tube with a calibre 
of about one-eighth of an inch, the shorter leg rather longer than the height 
of the flask, and the other leg four or five inches longer. I find the most con- 
venient way of purifying the syphon is to boil it, and in order to adapt it for 
packing into a saucepan, the glass tube is interrupted at intervals of about 
four inches with pieces of caoutchouc tubing, the shorter leg having one such 
india-rubber hinge, and the longer leg two. They are tied firmly on the tube 
with fine wire, silver wire being used for the shorter leg, where iron must rust. 
In the longer leg one of the caoutchouc junctions serves the further important 
purpose of enabling the assistant to control the flow through the syphon by 
compressing the india-rubber between the finger and thumb. A fourth piece 
* It is only in my more recent experiments that I have thus employed the cotton, but there can be 
no doubt that while it scarcely adds to the trouble of the process, it must materially increase its security. 
