362 



Journal of Agricultural Research 



Vol. IV, No. 4 



DUPLICATE SAMPLING WITH RETTGER AEROSCOPES SET UP IN TANDEM 



A series of 20 duplicate analyses was made with Rettger aeroscopes, 

 using 18-liter bottles as aspirators and filtering 15 liters of stable air 

 through each aeroscope. Five c. c. of sterile water (used instead of 

 physiological salt solution in order to avoid foaming) was used as the 

 filtering agent and i c. c. of the suspension was added to each Petri 

 plate. The aeroscopes were also set up in tandem connected by a 

 continuous glass tube. Table VIII gives the results of these analyses in 

 detail. 



Table VIII. — Duplicate counts made with Rettger aeroscopes in tandem couples 



Sample. 



Duplicate. 



Test No. 



1 



2 



3 



4 



5 



6 



7 



8 



9 



10 



II 



12 



13 



14 



15 



16 



17 



18 



19 



20 



Average 



Plate count 

 of first 

 tube. 



5 

 15 

 31 

 5 

 3 

 12 

 29 



71 



41 



171 



35 



52 



166 



51 

 132 

 108 



Plate count 



of second 



tube. 



0.6 

 •3 



II. o 



.6 



9-3 

 3-0 

 6.0 



S-o 

 2.6 



6.0 



Count per 

 liter. 



4 

 10 



3 



4 



24 



14 

 57 

 12 



17 

 55 

 17 

 44 

 36 



16 



Plate count 

 of first 

 tube. 



24 



42 

 3 

 4 



5 



17 

 8 



5 

 47 

 20 



57 



13 



116 



52 



50 



124 



129 



Plate count 



of second 



tube. 



0-3 



.6 



6.0 



•3 

 •3 

 •3 

 •3 



30 



2.8 



Coiuit per 

 liter. 



O 

 2 



4 

 14 



I 

 I 

 2 

 6 



3 



2 



16 



7 

 19 



4 

 39 

 17 

 17 

 41 

 43 



Many of the results show very few colonies per plate, especially where 

 the analyses were made when the cows were out of the stable and every- 

 thing was quiet. In this respect these figures are unsatisfactory, since 

 low figures greatly magnify contamination errors. The averages of 20 

 tests gave 48 colonies per plate for the first sample and 39 for the dupli- 

 cates. The second tubes of each tandem couple gave average counts of 

 6 (88.8 per cent efficiency) and 2.8 (93.3 per cent efficiency) colonies per 

 plate, respectively. 



These three series of comparative tests show clearly that caution must 

 be exercised in accepting the figures of any single comparison as correct. 

 On the other hand, the average of a series of comparative analyses prob- 

 ably gives reliable results. 



