170 DAIRY BACTERIOLOGY 



hand and shaking vigorously. For each tube a different portion 

 of the palm should be used, and, when completely wetted, the 

 hand should be carefully washed before proceeding any further. 

 In actual practice, the colour is added when all the samples have 

 been taken and the tubes are in position in the stand. After 

 placing in the water bath, the samples should be examined at 

 frequent intervals during the first twenty minutes, after which 

 they need only be examined every quarter or half hour. As 

 mentioned above, the author has found it best to employ the same 

 temperature as in the fermentation test ; in this way there is the 

 additional advantage that the two tests may be combined. The 

 combined reductase and fermenting test l gives information regard- 

 ing both the number and the nature of the organisms in the milk. 



The joint investigations of Barthel and the author 2 have shown 

 that it is possible by means of the reductase test to grade milk 

 and cream into four classes, as follows : 



Class 1. Good milk, not decolorised in five and a half hours, 

 containing, as a rule, less than | million bacteria per cubic centi- 

 metre 3 . 



Class 2. Milk of fair average quality, decolorised in less than 

 five and a half hours but not less than two hours, containing, as 

 a rule, | to 4 million bacteria per cubic centimetre. 



Class 3. Bad milk, decolorised in less than two hours, but not 

 less than twenty minutes, containing, as a rule, 4 to 20 million 

 bacteria per cubic centimetre. 



Class 4. Very bad milk, decolorised in twenty minutes or less, 

 containing, as a rule, over 20 million bacteria per cubic centimetre. 



If samples of retail milk from different dairies are to be com- 

 pared, they must, of course, be examined simultaneously ; thus 

 it would be unfair to sample one dairy on a cold morning and 

 another on a warm afternoon. Unpasteurised milk, as sold in 

 large towns, should retain its colour in the test for at least two 

 hours, and the pasteurised rmlk for at least five and a half 

 hours. Most of the milk retailed by the large dairies in Copen- 

 hagen fulfil these requirements 4 . The pasteurised milk supplied 



1 " Maelkeritidende," 1909, p. 359. 



2 " Milch wirtschaftliches Zentralblatt," 1912, No. 14. 



3 As the counts were found by the plating method, they were really 

 under-estimated ; milk bacteria usually occur in pairs and not infrequently 

 in long chains, or large clumps, which are not broken up on shaking, and in 

 such cases only one colony is obtained. The counts should at least be 

 doubled, and in some cases they should be trebled or quadrupled. 



4 Orla Jensen, " Maanedsskrift for Sundhedspleje," 1909, p. 239. The 

 translator has found the reductase and fermenting test to be of great use 

 in the control of milk as it arrives from the farms, and that the conditions 

 of treatment of the milk on various farms were found to correspond with 

 conclusions which had previously been drawn from the behaviour of the 

 samples in this test. 



