July IS, 191 s Methods of Bacterial Analyses of Air 359 



again after an interv^al of three hours. The detailed results will not be 

 given, since they were somewhat contradictory and bear upon another 

 line of investigation now in progress. In some cases there was a great 

 reduction in numbers of colonies appearing on the plates after the bac- 

 terial suspensions had stood for the three hours, in others little or no 

 reduction, and in some cases even an increase. There seemed to be 

 little difference between the action of the distilled water and that of the 

 physiological salt solution. 



One interesting result that appeared as the result of these comparative 

 tests was that the Rettger filters in this series of analyses caught nearly 

 though not quite as many bacteria as the sand filters. The exact averages 

 for the 26 tests were 142 colonies per liter for the Rettger aeroscope 

 as compared with 173 for the modified standard aeroscope. Of all 

 analyses given in this paper, 82 give a direct comparison between the 

 Rettger aeroscope and the modified standard aeroscope. Of these 82 

 comparisons, 49 show larger numbers of bacteria caught by the sand 

 filters, 17 show excesses for the water filter, and 16 give practically the 

 same figures for both. However, if we take only the last 54 of these 

 analyses, omitting the first comparisons, which were not properly carried 

 out in all their details, the showing is more favorable for the Rettger 

 aeroscopes. In these there were 22 excesses for the sand filter, 16 for the 

 liquid filter, and 16 where the results were practically identical. 



From these analyses it must be concluded that the Rettger aeroscope 

 is probably very nearly as efficient as the modified standard aeroscope in 

 the hands^of an experienced man. There are several things about it, 

 however, that are liable to cause trouble and others that make it less con- 

 venient to use. One already mentioned is the possible injurious effect 

 of the liquid used as a filter upon the bacteria before the material can be 

 plated, owing to absorbed gases, unfavorable osmotic action, or other 

 causes making it necessary to complete the plating as quickly as possible 

 after starting the analysis. This one objection makes it inadvisable to 

 use this type of aeroscope for analyses made at a distance from the 

 laboratory. A second difficulty is the need of great care to prevent bac- 

 teria from being held on the moist inner surface of the inlet tube. A 

 third is the necessity of steam sterilization, which occasionally loosens the 

 joints about the cork, causing inaccuracies. Since the modified form of 

 aeroscope met all of these difficulties and was cheaper, easier to operate, 

 less likely to break, and more adaptable to field work, it v/as decided to 

 use this in the investigation to v/hich the work recorded in this paper was 

 preliminary. 



COMPARISO.V OP RESULTS OBT.\I.\ED BY DUPLICATE ANALYSES MADE WITH THE SAME 



T\'PE OF FILTER 



Inasmuch as the conclusions drawn from data already presented are 

 mainly based upon the averages of from 25 to 30 comparisons, it maybe 

 instructive to find out whether such conclusions are justified or whether 



