154 The Principles of Vegetable- Oar dening 



Good 



Fair 



80 



80 



83 



80 



200 



200 



79 



75 



80 



80 



86 



80 



94 



90 



96 



90 



81 



80 



Eye . . . 

 Sunflower 

 Timothy . 

 Tobacco . 

 Vetches . 

 Wheat. . 



Good 

 . 95 

 . 94 

 . 96 

 . 78 

 . 96 

 . 95 



Ehubarb 

 Salsify . 



Spinach . . . . .. 



Squash 



Tomato 



Turnip, Flat . . . 

 Turnip, Eutabaga 

 Watermelon . . . 



Good 



Barley 93 



Broom Corn 92 



Buckwheat 94 



Clover 90 



Corn 92 



Grass Seeds 70 



Since variable results are obtained under any treat- 

 ment and from the same parcel of seeds, it follows 

 that some check must be employed in order to reach 

 reliable results. Two kinds of checks are open to 

 the investigator : (1) the selection of a representative 

 sample, and (2) duplicate trials. The greater the 

 number of seeds used in any test and the greater the 

 number of tests, the more reliable are the results. 



In choosing a sample, the contents of the whole 

 package should be thoroughly mixed, particularly if the 

 package has been shipped and the heavier seeds have 

 settled. The seeds for trial should then be drawn 

 from various parts of the package, or from its center 

 if the package is small, and the various lots mixed 

 and the sample for testing taken from the mixture. 



