Comparisons of Tests of Milk Samples taken Conventionally and with an Automatic In-Line Sampler ¹

Samples of manufacturing bulk milk were collected from 102 producers by the conventional procedure and were randomly coded. At least 3 h later, samples were collected by haulers from the same tanks with in-line samplers (ISOLOK MS5) mounted on the truck. Overall weighted mean fat tests for producers on nine routes were 3.669% for conventional and 3.644% for in-line samples. The overall weighted mean of load samples taken at the plants was 3.655%. Unweighted means of the producer samples taken by the conventional and in-line methods were 3.781% and 3.752%, respectively. The difference between the means, 0.029%, was highly significant (P<0.001). There was a significant difference (P<0.05) in fat content of samples taken conventionally after 5 and 8 min of agitation. Carryover of bacteria from tanks with high counts to samples with low counts was not detectable by direct microscopic counts in field studies. However, in laboratory studies with an emptied hose and a few seconds of delay before sampling small carryover occurred. Significance of the carryover effect depends on volume of milk sampled and whether bacterial content is marginal.