METHODS OF INVESTIGATING BACTERIA. 805 



point a was heated and then broken off. A small 

 quantity, 1 to 3 litres, of air was drawn through the 

 fluid and the point a was sealed, the plug c, composed 

 of spun glass, and which might have taken up some of 

 the germs, was blown back into the fluid, and the ap- 

 paratus was placed in an incubator. In each of these 

 experiments 20 to 50 vessels were employed ; after some 

 time it was ascertained in how many of these muddi- 

 ness had occurred as the result of the development of 

 bacteria, and if this did not occur in all the vessels, but 

 only in a small proportion of them, it was assumed that 

 in each of the muddy vessels only one germ had entered. 

 Hence the number of the vessels in which develop- 

 ment occurred gave also the number of the germs which 

 were present in the quantity of air aspirated through 

 all the vessels. If none of the tubes became muddy, or 

 if, on the contrary, they all became turbid, the experi- 

 ment was repeated with larger or smaller quantities of 

 air ; if all the tubes had become turbid it was only 

 possible to obtain the approximate minimal number of 

 the organisms present, for in each case the muddiness 

 might have been due not only to one but even to two, or 

 ten, or more germs. 



As is quite evident, the utility of this method rests on Sources of 

 the assumption that the germs are very equally dis- me 

 tributed in the air, and that they do not exist in masses. 

 All other observations, however, show that this assump- 

 tion is incorrect. Direct microscopical observations 

 demonstrate that there are numerous collections of 

 bacteria among the germs present in the air, and that 

 they are by no means equally distributed through- 

 out the air. Besides, the whole method is extremely 

 troublesome, and does not permit any sufficient varia- 

 tion of the nutrient media. 



Hesse has attempted to utilise the solid nutrient sub- Hesse's 

 strata for the investigation of air. A glass tube, 70 cm. m 

 in length and 3' 5 ccm. in width, is filled with 50 ccm. of 

 nutrient jelly in such away that the inner wall is lined with 

 the material, and that a thicker layer of it is present at 

 the lower part of the tube. One end of the tube is closed 



