356 Journal of Agricultural Research voi. iv. no. 4 



A break has been made in Table IV because of a slight change which 

 was made in the method of plating the bacterial suspensions from the 

 Rettger tubes. In the first 15 analyses the aeroscope arm through which 

 the air enters was washed out once, just before the bacterial suspension 

 was used for plating, by drawing the suspension up into this arm by suc- 

 tion and then releasing it. This was done because it was realized that 

 there was danger that some bacteria would be lost by clinging to the 

 moist inner surface of this tube. As the counts were made, it became 

 evident that the number of colonies appearing per liter from the 

 Rettger aeroscope was less than the number appearing per liter from 

 the modified sand aeroscope. For this reason, beginning with test No. 

 16 the rinsing of the entrance arm to the Rettger aeroscope was done 

 more thoroughly (several times). By this means the number of colonies 

 appearing on the Rettger plates was increased, showing that the sur- 

 mised effect of this moist tube was probably true. This effect of the 

 long entrance tube in reducing the number of bacteria in the water filter 

 is also realized by Rettger, for he states (22, p. 467) that one may expect as 

 high as a 15 per cent error in this way. For this reason he recommends 

 drawing the tube out of the aeroscope (after the steam sterilization and 

 before use) and flaming it in order to make sure that the tube will be dry 

 when used. This was not done in the present series of tests, because it 

 was felt that this manipulation after sterilization introduced too great a 

 chance of accidental contaminations. 



The results secured before and after the change of procedure noted 

 were as follows: When the entrance tube was rinsed but once, the Rettger 

 aeroscope gave an average of 25 and the modified standard aeroscope 

 gave 121 colonies per liter. After the rinsing was done more thoroughly, 

 there was still a large discrepancy in results, the Rettger filter giving an 

 average of 36 and the modified sand filter an average of 97 per liter. 



In the first 1 5 tests the average counts obtained from the second tube 

 of each tandem pair was 2.3 colonies per plate for the modified standard 

 aeroscope and 5.8 per plate in the case of the Rettger aeroscope. On 

 the basis of these numbers, the percentage efficiencies calculated for the 

 two aeroscopes are 97.7 and 87.3, respectively. The low percentage 

 efficiency obtained with the Rettger aeroscope is somewhat misleading, 

 owing to the fact that in this case the total numbers obtained by both 

 aeroscopes of th^ tandem pair are low. In the last nine tests the per- 

 centage efficiencies were 98.8 for the modified standard aeroscope and 

 97.1 for the Rettger aeroscope. 



