December, 1921] results of seed tests, 1921. 21 



samples submitted in excess of the number tested free of charge. 



Sect. 12. Chapter 61, Laws of 1909, as amended by chap- 

 ter 89, Laws of 1915, and all acts or parts of acts inconsistent 

 herewith are hereby repealed. 



Sect. 13. This act shall take effect September 1, 1921. 



Approved April 8, 1921. 



Rules and Regulations Relative to the Provisions and 



Enforcement of the Law Governing the 



Sale of Agricultural Seeds 



(CHAPTER 80— SESSION LAWS OF 1921.) 



1. Definition of word "approximate." By approximate 

 percentage of purity (Sec. 2) is meant the actual percentage of 

 purity within certain prescribed limits of variation. The varia- 

 tion or tolerance allowed by the term "approximate" is known 

 as "double tolerance" and is computed by the following formula: 



T = .2 ( .1 H ^ ) where "T" represents the variation to be 



allowed; ".2" an arbitrary constant used; and "P" the percent- 

 age of purity found. For example if a sample of seed is found 

 to be 94.8% pure the variation allowed according to the formula 

 would be 1.24, and a statement of 96% purity would be con- 

 sidered "approximately" correct. 



A detailed explanation of purity variations and a table of 

 tolerance for them will be found on pp. 19 and 20 of the Pro- 

 ceedings of the Association of Official Seed Analysts of North 

 America for 1917. 



By approximate percentage of germination (Sec. 2) ^ is 

 meant the actual germination within the limits of variation 

 allowed by the Rules for Testing adopted by the Association of 

 Official Seed Analysts of North America. At the present time 

 these variations are as follows: 



Stated Percentage of Germination I Percentage of Variation 



1 Allowed. 



90 or over I 6 



80 or over and less than 90 1 7 



70 or over and less than 80 | 8 



60 or over and less than 70 | 9 



40 or over and less than 60 | 10 



For example, if a sample of seed was found to germinate 

 91% and was certified to germinate 96% the latter ^statement 

 would be considered "approximately" correct because it is with- 

 in the percentage of variation allowed. 



2. Taking of samples. 



Failure to secure representative samples is one of the 

 most common causes of variation in tests of seed. The fol- 

 lowing directions for sampling are given. 



(1.) The minimum weight of seed for tests should be 

 approximately: 



(a) one ounce of grass seed of any kind or of white 

 and alsike clover; 



(b) two ounces of red and crimson clover, alfalfa, mil- 

 let, fliax, or seed of like size. 



(c) one-half pound of cereals or seed of like size. 



