2 N. H. EXPERIMENT STATION [Bulletin 197 



marked ^•'Satisfactory/' as explained in a later paragraph, our 

 policy in every case has been to retest. If the second test showed 

 any considerable variation from the first, a third would be made; 

 otherwise, the average of the two tests would be reported. In this 

 way an endeavor has been made to eliminate any gross errors, or 

 results of abnormal conditions. 



For several years past our reports have been marked "Above" 

 when our findings have been better than the guarantee, "Satisfac- 

 tory" when our findings came within the variations previously sug- 

 gested, and "Below" when the seed could not be classified under 

 either of the two preceding terms. The first two words appear to 

 serve their purpose, but the last one is hardly explicit enough. 

 The dealer guaranteeing seed at 90% germination, for example, 

 which tests 85 per cent and would have to be marked "Below" is 

 hardly in the same class with the dealer who makes a similar guar- 

 antee for seed which only germinates 25 per cent. For this reason, 

 seeds which have previously been marked "Below" will be accompa- 

 nied in this report by actual figures for germination or purity as 

 the case may be. 



