354 



Journal of Agricultural Research 



Vol. IV. No. 4 



being 132 (13.2 colonies per plate) and the lowest o. Undoubtedly a 

 large part, if not all, of the bacteria present were really in the sand or 

 clinging to the walls of the aeroscope and were scraped out of the tube 

 with the cotton plug. Even if they were all present in the cotton, they 

 represent less than i per cent of the total number of bacteria caught. 



All of the analyses thus far described were made by the writer. Since 

 it was desirable to know whether another worker would get equally good 

 results with this modified form of aeroscope and also whether the sand 

 filter was equally efficient when a more rapid current of air was passed 

 through it, Mr. Kulp, who had never before had experience with bacterial 

 air analyses, undertook a new series of analyses. In this series of 22 

 tests, the air was aspirated at the rate of i liter per minute or even 

 faster in some cases (7 liters in five minutes), while all of the previous 

 analyses had been made with an aspiration rate of i liter in two minutes. 

 Connections were sealed with paraffin, as before. The material was 

 plated on both agar and lactose agar, but only the results of the former 

 will be discussed here. The samples were plated in triplicate, as in the 

 preceding series. The results of this series of tests are given in Table III. 



Table III. — Efficiency tests of the modified standard aeroscope 

 [Aeroscopes set up in tandem. Aspiration more rapid than in previous tests] 



Test No. 



3 

 4 



5 

 6 



7 

 8 



9 

 10 

 II 

 12 

 13 



14 



15 



16 



17 



18 



19 



20 



21 



22 



23 



A\'erage 



Plate 

 count in 

 first tube. 



56 



93 



75 

 60 



43 

 15 

 66 



43 

 30 

 43 



76 



Plate 



count in 



second 



tube. 



Count 

 per liter. 



80 

 107 



86 

 61 



94 

 61 



43 

 61 



105 



From this table it will be seen that the first tube of the modified form 

 of aeroscope gave an average of 76 colonies per plate (96.3 per cent), 

 while the second tube gave 3 colonies per plate {t,.j per cent). The 

 cotton plugs were plated as before, resulting in the development of 36 

 colonies per plug (3.6 colonies per plate). Since the number of colonies 

 that appeared on the plates made from the second tube was the same as 

 in the previous series (Table II, second series), it was felt that the lower 

 percentage of efficiency in this case was largely the result of the fact that 

 fewer bacteria were present per liter of air. 



