SEED TESTS. 



Made September 1, 1916, to September 1, 1917. 



The provisions of the Pure Seed Law enacted by the General 

 Court in 1909 require the publication annually of a bulletin 

 showing the results of all seed tests made officially during the 

 previous year. The administration of this law is in the hands 

 of the State Commissioner of Agriculture, who has appointed 

 the writer, Agronomist of the Experiment Station, as his regular 

 agent for making all tests and analyses in this state. 



During the year ending September 1, 1917, one hundred and 

 three samples of seed were collected and sent in for analysis 

 by an authorized representative of the Commissioner of Agri- 

 culture. These samples were secured from twenty-eight dealers 

 in twenty-one different towns, representing each of the ten 

 counties of the state. The samples included 7 of alsike clover, 

 13 of red clover, 4 of white clover, 4 of barley, 8 of corn, 14 of 

 millet, 3 of oats, 12 of redtojD, 26 of timothy, and 9 of miscel- 

 laneous seeds. 



In the testing and analyzing of seeds, just as in the chemical 

 analysis of fertilizers, it should be remembered that there are 

 certain factors which make exact duplication of results impos- 

 sible. There are always slight variations in drawing a sample 

 of seed, in sampling the sample, in weighing small fractional 

 parts of it, in the counting and identification of numerous 

 seeds, etc. ; there may be also fluctuations in the temperature 

 and humidity of the germinating chamber which may affect the 

 percentage of germination. Methods of analysis may also vary 

 somewhat. For these reasons a slight variation from the guar- 

 antee or standard, or from a previous analysis, should not be 

 considered of consequence. 



In publishing the results of this year's samples we have again 

 made use of a table of "tolerance of variation" for purity which 

 has been prepared by one of the leading seed laboratories of the 

 country, and which seems to afford a very satisfactory method 

 of allowing or tolerating a reasonable variation in purity tests. 

 The percentage variation for any given sample is determined 

 bv the following formula : 



2(.l-fl00— P) 



T= 



10 

 in which "T" is the variation tolerated and "P" is the per- 

 centage of purity found. This formula applies only to seed 

 having a purity of 50% or greater. 



