12 THE VITALITY AND GERMINATION OF SEEDS. 



folds, to dip into water in the pan. This should never be permitted, 

 for in that way seeds are kept too moist, especially near the ends of 

 the folds. Likewise such methods give an opportunit}' for the trans- 

 mission of dissolved copper and a resulting- injury to the seeds. For 

 this same reason the strips of cloth should be made sufficiently narrow 

 not to come into contact with the sides of the pan. 



Much better results are obtained if the seeds, before being placed 

 in the germinator, are soaked in water for several hours, the length 

 of time depending on the power of absorption of the seeds. In these 

 experiments the seeds were always soaked in distilled water for twelve 

 or fifteen hours before transferring them to the germinator. This 

 preliminar}' soaking gives a more speedy germination, which is always 

 advantageous, especially in making comparative germination tests. 

 In order to supply the requisite amount of moisture for subsequent 

 growth, the cloths were first uniforml}^ and completely" wet with dis- 

 tilled water; moreover much care was taken to see that there was only 

 a very small quantity of water in the bottom of the pan. In case of 

 seeds that germinate readih% such as cabbage, lettuce, and onion, it is 

 necessary that all surface water be removed from the bottom of the 

 germinator if good results are desired. The pan then being covered 

 with a glass plate, it is seldom necessary" to increase the amount of 

 moisture, for seeds when once soaked need only to be kept slightly 

 moist and not wet, as must necessarily be true if the ends of the cloths 

 or bottoms of the folds dip into the water. After soaking, the water 

 in the seeds and cloths is ample for the completion of most germina- 

 tion tests. However, in an occasional test the seeds may become 

 slightly dry, which happens when the cover is kept off the pan for a 

 considerable time while counting germinated seeds. In such cases the 

 remedy is to pour a small quantit}" of water in the bottom of the pan, 

 or in extreme cases to moisten the folds with a fine spray. 



If the above modifications be adopted and the necessary precautions 

 taken, many of the objections frequently made to the Geneva tester 

 will be removed and the difficulties will be overcome; at least it is a 

 most excellent method of testing seeds where comparative results are 

 especially desired. It must also be borne in mind that the Canton flan- 

 nel (which is generally used in making the pockets) should alwa3^s be 

 of the best grade and should never be used a second time without being 

 thoroughly cleaned and sterilized. 



In selecting samples for germination the impurities and the imma- 

 ture seeds were first removed. The samples for test were then made 

 up of the remaining large and small seed. For the most part 200 

 seeds were taken for a test, but with the larger seeds — corn, pea, bean, 

 and watermelon — 100 seeds were usually used. In all cases where any 

 irregularity was apparent, tests were repeated. The controls are 

 based on the results of several duplicate tests. 



