AGRICULTURAL EXPERIMENT STATION. 17 



In Table V a comparison is made between tlie manufacturer's 

 minimum guarantees and the actual composition of both the 

 samples sent to the Station by the manufacturer and those selected 

 in the open market by the Station. 



This comparison may be summarized as follows : 



(1.) Fifty-one brands are involved in this comparison. 



(2.) The manufacturer's samples were as good or better than 

 the minimum guarantee thirty-four times for the nitrogen, forty-one 

 times for the phosphoric acid and forty-six times for the potash, 

 while the manufacturer's samples were poorer than the minimum 

 guarantee, seventeen times for the nitrogen, ten times for the 

 phosphoric acid and six times for the potash. 



(3.) The Station samples were better than the manufacturer's 

 samples ten times for the Ditrogen, seventeen times for the 

 phosphoric acid and thirteen times for the potash, while the Station 

 samples were poorer than the manufacturer's samples thirty-two 

 times for the nitrogen, thirty-two times for the phosphoric acid 

 and thirty-six times for the potash. 



(4.) The actual average differences between the percentages 

 of nitrogen, available phosphoric acid and potash in the manu- 

 facturer's and Station samples were in favor of the manufacturer's 

 samples to the following extent : Nitrogen .13 per cent., available 

 phosphoinc acid .23 per cent, and potash .30 per cent. This 

 means that on the average the samples selected by the Station 

 were about 5 per cent, poorer in nitrogen, 2 per cent, poorer in 

 phosphoric acid and 10 per cent, poorer in potash than those sent 

 to the Station by the manufacturers. 



The Station samples, on the other hand, are on the average 

 equal to the minimum guarantees. 



(5.) It is fair to remark in view of the foregoing statements 

 that the manufacturer's samples have much more closely repre- 

 sented the goods in the market in nitrogen and phosphoric acid 

 than in potash. It is not easy to explain why in thirty-six cases 

 out of fifty-one the Station samples selected by the Station in the 

 markets should fall so much below the manufacturer's sample in 

 the percentage of potash soluble in water. This fact is not 

 accounted for, as some might suggest, by the errors of sampling 

 small lots of goods which do not properly represent the entire 

 bulk, because in that case the percentages would be too large as 

 often as too small. Moreover, if it were a question of sampling, 

 the nitrogen would be affected to a like extent. Some years ago 



