1900] MICROSCOPICAL JOURNAL. 35 



maining constant. The plates were incubated in a moist 

 chamber, in a closed chamber without moisture, in the 

 incubator (or the ice chest), and in a desiccator. For the 

 moist chamber a desiccator-jar was used, the sulphuric 

 acid being replaced by water. The closed chamber was 

 a desiccator-jar without sulphuric acid or water. The jars 

 were large euough to hold five plates, and in all cases the 

 figures representing the numbers of bacteria found were 

 obtained from the averages of the five counts. 



The largest numbers of bacteria were always obtained 

 from the plates that developed in the moist chamber, and 

 the smallest numbers were always obtained from the 

 plates that were kept in the desiccator. If the number 

 that developed in the moist atmosphere be assumed to 

 represent the maximum number obtainable by the meth- 

 od, it follows that on 78 per cent of the bacteria develop- 

 ed in the desiccator, that 85 per cent developed in the 

 closed chamber, and that the numbers that developed in 

 the ordinary incubator and iu the ice chest varied from 

 75 per cent to 98 per cent. There was a striking uniform- 

 ity in the percentages which the numbers in the desicca- 

 tor were of the numbers in the moist chamber and this 

 was due, no doubt, to the constantatmospheric conditions 

 in the two jars. In the closed chamber also the percent- 

 ages were quite constant. In the ice chest and in the in- 

 cubator the percentages were more variable. The atmos- 

 phere of the ice chest was ordinarily more moist than that 

 of the incubator, but was seldom saturated on account of 

 ventilation. The amount of moisture in the air of the in- 

 cubator was generally greater than that of the atmos- 

 phere of the laboratory, and. varied more or less with the 

 ventilation, the number of plates kept in the incubator, 

 etc. In one Series the relative humidity of the atmos- 

 phere in the incubator wasestimated at about 60 per cent. 

 In another Series the humidity was determined by hourly 

 readings of a psychrometer and was found to vary be- 



