APPENDIX. 165 



the bottom of the pail, it is possible to get a sample miicli less 

 rich in fat than the top would yield. If there is more than one 

 pail of cream, a portion should be taken from each and every lot 

 weighed out. If a tube full from every lot more than fills the 

 sample bottle, then all the portions drawn should be mixed in a 

 dish large enough to hold them and the bottle filled from the 

 mixture. In any case enough should be taken to fill the sample 

 bottle to prevent churning on the road. 



Preservative. — For this purpose we would recommend bichromate 

 of potash. After the sampling bottles are thoroughly cleaned 

 with hot water and washing soda, a small amount, about what can 

 be held on one-quarter inch of the blade of a penknife, of the 

 finely powdered bichromate of potash should be put in each bottle 

 before starting out to collect cream, then if the cream is perfectly 

 sweet and well shaken up after being put in the bottle, it will keep 

 sweet two weeks if kept in a cool place. Cream that is sour should 

 not be sampled, as it is impossible for a collector to get a fair 

 sample of it in any reasonable length of time. If it has become 

 thick it cannot be mixed by the collector so it will be uniform, and 

 cannot be sampled with the tube. Creamery managers should in- 

 sist that their patrons keep tbeir cream sweet until it is taken by 

 the collector. This is essential not only to correct sampling, but 

 to making a good quality of butter. 



Composite Sample. — The composite sample is made up from the 

 small samples taken by the collector and is the one from which the 

 portion is taken for the test. Half-pint fruit jars are good recep- 

 tacles in which to put these samples, and each one should be num- 

 bered with the patron's number, the same as the small bottles used 

 by the collector. 



The small samples are taken every time cream is collected 

 according to the directions previously given, and as soon as they 

 arrive at the factory they are emptied into the fruit jars having 

 corresponding numbers. The jar should be closed tightly to pre- 

 vent evaporation. At the end of two weeks, if one chooses to test 

 as often as that, these accumulated small samples constitute the 

 composite sample and the per cent of butter fat found in this 

 sample will be the average per cent in all the cream furnished by 

 the patron having that number for that period. 



In winter, or if the composite sample be kept in an ice chest 

 with the preservative, it could be kept four weeks as easily as two, 

 thereby reducing the work of testing one-half. 



