DETERMINATION OF AMOUNT OF DIRT IN MILK. 289 



and then by being placed in a desiccator for a day or two. 

 It is then to be weighed upon the chemical balance. After 

 weighing it should be replaced in the desiccator for another 

 twenty- four hours and again weighed. If the two weighings 

 agree, the weight may be taken as the proper weight, but if 

 the second weight is less than the first, it indicates an insuffi- 

 cient drying of the filter paper, and it must be placed again in 

 the desiccator until two successive weighings agree. The 

 weight of the dried filth will be obtained by subtracting the 

 original weight of the filter paper from the final weight. 



Centrifugal Method. This has the advantage of being 

 much more rapid and probably fully as accurate as the first. 

 In most respects it is identical. Filter papers must be dried 

 and weighed as before, and all of the methods of procedure 

 are the same except the method of separating the sediment 

 from the milk. For this purpose a measured sample of the 

 milk is placed in one of the vials of a centrifuge and rapidly 

 rotated. The rapidity of rotation should be about 2,000- 

 3,000 times per minute. After the dirt has thus been sedi- 

 mented the supernatant milk is pipetted off in the usual way, 

 and the sediment, after being washed as before, is collected 

 upon filters. The advantage of this method is its much 

 greater rapidity, and the disadvantage is the fact that it is 

 necessary to use smaller amounts of milk, since the centri- 

 fuges in use are not adapted for the treatment of large 

 quantities. With this method, therefore, it is especially im- 

 portant to use the greatest precaution in selecting the sample, 

 for any irregularities in the sampling will produce very great 

 errors in the result. In order that the sampling may be accu- 

 rate it is, of course, necessary that the milk should be thor- 

 oughly stirred and the sample taken should be as nearly as 

 possible an average sample of the whole milk. In all methods 

 of estimating the dirt in milk it is necessary to be especially 

 careful in the matter of selecting a sample. 



'9 



