3 N. H. EXl'EEIMENT STATION. 



Germination in Per Cent. Allowable Variation 



90 and over 6 



80 or over, but less than 90 7 



70 or over, but less than 80 8 



60 or over, but less than 70 9 



Less than 60 10 



Purity testing work, in wliich the sample is separated 

 into pure seeds, other crop seeds, weed seeds, and inert 

 material, also involves some variation due to normal and 

 unavoidable small errors tending always to prevent any 

 two samples or separations from being exactly alike. 

 The variations in this case, however, are much less than 

 are to be expected in germination and are provided for 

 by a formula. This purity tolerance formula, so-called, 

 considers the sample as made up always of two parts, 

 (1) the element, or component part, being considered 

 and (2) the rest of the sample. The tolerance, then, in 

 per cent, allowed for each component part shall be : two 

 tenths of one per cent, plus 20 per cent, of the lesser 

 part. This formula to be applied to the analysis that is 

 guaranteed. 



If a sample is guaranteed to germinate 90 per cent. 

 and to be composed of 90 per cent, pure seed, it can 

 average in our laboratory tests, according to this inter- 

 pretation, between 84 and 96 per cent, germination and 

 between 87.82 and 92.18 per cent, of pure seed and still 

 be "satisfactory." If we have found the sample better 

 than the highest figure for variation allowed, it is 

 marked "above" in this bulletin, if lower than the low- 

 est figure for variation allowed, the actual figure is re- 

 ported. 



From the summary table at the end it will be ob- 

 served that large proportion of the seeds collected this 

 year is satisfactory in both purity and germination, a 

 small number is below, and a still smaller group in each 

 case is above the variations allowed. In actual figures 



