SEED TESTING 113 



Instead of the window-box, paper flower-pots may be used 

 to great advantage for this seed testing. Fill a pot with 

 good loamy soil to within an inch of the top, then scatter 

 evenly over the surface twenty-five or fifty or one hundred 

 of the seeds to be tested, cover these with a half inch of sifted 

 garden soil, water carefully, and watch the seedlings in the 

 way described above. 



In general, the highest percentage of good seeds will be 

 found from the sprouting test, the next highest will be found 

 from the seedling test, and both of these will be likely to be 

 higher than the actual results that can be obtained in the 

 garden out-doors. In the latter the conditions are generally 

 less favorable to the development of the seedling than are 

 the conditions of the in-door tests. 



It is much more important to make tests of some seeds 

 than it is of others. Two sorts of seeds in which it is well 

 worth while to make such tests are tomatoes and onions. In 

 case of the tomatoes, cheap seeds will often give a compara- 

 tively small percentage of good plants, and it is most im- 

 portant that the tests should be made so early that if the 

 results are not favorable other seeds may be obtained in time 

 for planting. For example, a great many poor onion seeds 

 are sold, and it is very important that the percentage of 

 germination of a given lot be determined before the main 

 sowing is made. In case the seeds should prove to have a 

 low percentage of germination, it will be desirable, of course, 

 to sow the seed more thickly, or very often it will be desirable 

 to discard it altogether and substitute a better quality of 

 seed. 



In case of a few plants, however, older seed is preferred. 

 For example, when cucumber seed is ten years or more old, 



