July 15. 191S Methods of Bacterial Analyses of Air 349 



STUDY OF SAND FILTERS IN TANDEM 



A series of tests was made, using a sand-filtration method similar in 

 all respects to the standard method except that the rubber stopper with 

 the small tube was omitted from the top of the aeroscope. Two aero- 

 scopes were set up in tandem — i. e., connected end to end by means of a 

 glass tube and two rubber stoppers. In these tests, 5 liters of air were 

 drawn through the filters. No attempt was made to seal the connections 

 leading to the aspirator bottle, but they were made carefully in an 

 attempt to avoid leakage. The plating was done as described on page 346, 

 and in such a way that each plate represented i liter of air. In this series 

 of 28 tests the average number of colonies developing per liter was 59 in 

 the first tube (75.9 per cent) and 18.7 in the second tube (24.1 per cent). 

 If three cases, where leakage about the connections between the two 

 aeroscopes evidently occurred, are left out of the calculation, 25 tests 

 gave an average of 50 colonies developing per liter for the first tube 

 (91.4 per cent) and 4.9 per liter for the second tube (8.6 per cent). Later 

 results indicated that slight leakage about the apparently tight stoppers 

 may have influenced even the latter results. 



One of the difficulties constantly encountered in air-filtration work is 

 leakage. When this takes place at some point between the filter and 

 the aspirator, the result is that some of the air does not pass through the 

 filter, and therefore the results are low. When, as is probable in the 

 three cases mentioned above, the leakage takes place between the two 

 aeroscopes set up in tandem, the result is that the air goes through only 

 the second filter, making the percentage efficiency of the filter appear 

 lower than it really is. 



Another possible condition may have had a marked influence on these 

 results. At times, when sand-filter tubes are sterilized in the autoclave, 

 the sand sticks together to such an extent that cracks appear in it. 

 Such cracked filters are obviously not capable of catching all of the 

 bacteria because of the open pores or lanes through which they may 

 pass. In all of the tests of this series, however, as in all later series, such 

 obviously leaky filters were never used, so that the results given all favor 

 the standard aeroscope to that extent. 



TESTS OF RETTGER AEROSCOPES IN TANDEM 



A series of 33 tests was run with the Rettger aeroscopes in tandem in 

 order to determine their filtering efficiency. In this work again the 

 difficulty in securing absolutely tight joints was not properly appreciated 

 at first, and rubber connections were used between the two aeroscopes. 

 After sterilization in steam these rubber connections always had a ten- 

 dency to loosen, and while they were always examined to make sure that 

 they were tight, yet the results indicate that leakage did occur in three 

 cases at least. Seven liters of stable air were drawn through each aero- 

 scope. The aeroscopes contained 5 c. c. of physiological salt solution. One 

 c. c. of the resulting suspension was used in making the plates, agar being 

 92315°— 15 6 



