MICROORGANISMS AND FERMENTATIONS 21 



number of microorganisms generally takes place during transit. 

 Bacterial multiplication may, however, be avoided to a con- 

 siderable extent if the samples are packed in ice. For counts of 

 water samples the peptone gelatine recommended above may 

 be used, modified as follows : the sugar is omitted, and the 

 neutralised medium is treated with 15 c.c. of 10 per cent, soda 

 solution per litre. In these, as in all bacterial counts, it is im- 

 possible to get all the germs present to grow on the same medium ; 

 as many typical water bacteria, for example, the thread-forming 

 and sulphur organisms, will not grow on gelatine at all, the counts 

 must not be regarded as representing the actual numbers of 

 organisms present, though they furnish useful indications as to 

 the relative purity of samples if the same working conditions are 

 adhered to throughout. It must also be remembered that long 

 chains of bacteria or large pieces of mould mycelia only yield 

 single colonies. The method is thus subject to considerable 

 error, a fact which may easily be demonstrated by comparing the 

 results with those obtained by direct microscopical counts. 

 According to Barthel's investigations, 2 to 200 times as many 

 organisms are found by direct counts as by plating according to 

 Skar's method 1 . Pasteurised milk cannot be examined by the 

 direct method as there are no means of distinguishing by micro- 

 scopic inspection between living and dead bacteria 2 . 



Furthermore, one cannot be certain that any particular colony 

 may not have arisen from several different species which may 

 have adhered together or become intertwined in the original 

 liquid. If, therefore, a pure culture is desired, it will be necessary 

 to sow plates from a well-isolated colony, and only when the 

 resulting colonies are found to consist exclusively of the desired 

 organism is it possible to be sure that the individual colonies are 

 pure. It is still safer to make a single cell the starting point, a 

 method which is described under the next heading. 



Pure cultures are best preserved in Freudenreich flasks on 

 nutrient agar. As a rule stab cultures are made by piercing the 

 medium with a platinum wire on the point of which a trace of the 

 culture has been picked up. Streak cultures are made of organisms 

 which require free access to air ; the agar is allowed to solidify 

 in a slanting position so as to expose a large surface, and the 

 infected wire is drawn lightly across the surface. In order to 

 protect stab cultures of anaerobic bacteria from the air, the 

 cotton wool plug is flamed and pushed down the test tube till it 



1 " Milcliwirtschaftliches Zeiitralblatt," 1912, p. 455. 



2 H. W. Conn (U.S. Public Health Keport, 1915, No. 295) gives some 

 valuable information concerning the limits of accuracy in^the bacterio- 

 logical analysis of milk. 



