EXPERIMENT STATION. 17 



In case of a fine, uniform and moist or coherent article, a butter- 

 tryer or a tin tube, like a dipper handle, put well down into the 

 packages, in a good number of places will give a fair sample with 

 great ease. With dry, coarse articles, such as ground bone, there 

 is likely to be a separation of coarse and fine parts on handling. 

 Moist articles put up in bags or common barrels may become dry 

 on the outside. It is in these cases absolutely necessary to mix 

 thoi-ougbly the coarse and fine, the dry and the moist portions 

 before sampling. Otherwise the analysis will certainly misrepre- 

 sent the article whose value it is intended to fix. 



The quantity sent should not be too small. When the material 

 is fine and uniform, and has been carefully sampled, a pint may be 

 enough, but otherwise and especially in the case of ground bone, 

 which must be mechanically analyzed, the same should not be 

 lesH than one quart. 



It is also important that samples for analysis should be taken at 

 the time when the fertilizer is purchased, and immediately dis- 

 patched to the Station. Moist fish, blood or cotton seed will soon 

 decompose and lose ammonia, if bottled and kept in a warm place. 

 Superphosphates containing much nitrogen will sufi'er reversion of 

 their soluble phosphoric acid under similar circumstances. Most; 

 of the moist fertilizers will lose water unless tightly bottled, but 

 some of the grades of potash salts will gather moistui-e from the 

 air and become a slumpy mass if not thoroughly protected. 



These changes in the composition of a sample not suitably pre- 

 served must invalidate any conclusions from its analysis, and 

 work serious injustice either to the manufacturer or to the con- 

 sumer. 



It doubtless often happens that a purchaser on laying in a stock 

 of fertilizers decides that he will not then trouble the Station to 

 analyze the goods he has obtained, but will set aside samples 

 which he can send for examination in case the crops report 

 adversely as to their quality. It is always better to send all 

 samples at once to the Station where they can be directly analyzed 

 or so prepared that they will keep without chemical change. 



With the Instructions for Sampling, the Station furnishes a 

 blank Form for Description of Samples, a copy of which is here 

 given. 



