TESTING SEEDS AT HOME. 177 



of loss in tliis as in most grass seeds, but. il should not exceed 5 bush- 

 els in 10. Here is a clear loss of 4 bushels out of every 10 bought, 

 which, at $1.65 per bushel, is worth considering. The normal waste 

 in orchard grass seed is 1 bushel in 5, but the sample tested contained 

 almost 3-j- bushels of worthless seed out of 5. At present orchard 

 grass brings about $2 per bushel. This makes a net loss of about $7 

 on a purchase of 5 bushels of seed. It is unnecessary to give other 

 examples of the loss which farmers suffer by purchasing poor seed. 

 The table affords ample illustration. 



METHODS OF TESTING SEEDS. 



Many seedsmen and a few farmers test their seeds. The method 

 generally followed is to throw a handful of seed into a box full of 

 earth, and decide by the way it comes up whether the seed is good. 

 This is better than no testing at all, but it is impossible to get accu- 

 rate results in this manner if the seeds used are not counted. 



Another method is to make a shallow trench in sand, scatter in the 

 seeds as thickly as is recommended for the variety, and wet with warm 

 water. The seeds germinate rapidly, and the merit of the sample is 

 judged by the stand in the row. When the seeds are not counted, no 

 accuracy is possible. Besides, it is well known that the amount of 

 seed thought necessary per running foot of drill, or per acre, is from 

 two to four times as much as would be required if the seeds used had 

 a high vitality. 



Some people think that if seeds are thrown into water the good ones 

 will sink and the dead seeds will float, but this notion is not sup- 

 ported by facts. When seeds float it is often because an air bubble 

 has become attached to them or because they have not become wet 

 all over the surface. Several experiments were made to test the 

 germination of seeds that sink and those that float. Wheat was 

 used in one set of experiments, and the average of all tests showed a 

 germination of 68.3 per cent for the sunken seeds and 72 per cent for 

 those that floated. In another set of experiments lentil was used, and 

 it was found that 75.4 per cent of the sunken seeds and 86.7 per cent 

 of those that floated germinated. 



The germination of seeds depends 011 a proper supply of heat and 

 moisture. For accuracy in testing, darkness is also essential. Seeds 

 will germinate through a considerable range of temperatures, but 

 the number of germinating seeds decreases as we depart from the 

 optimum, or most favorable, temperature. If seeds are subjected to 

 temperatures higher or lower than the optimum, germination will 

 proceed more slowly, and when either extreme is passed it will cease. 

 All seeds do not have the same temperature limit. Seeds of tropical 

 plants need more heat to germinate than those from plants growing in 

 northern latitudes or on high altitudes. Certain seeds have been 

 known to germinate upon ice, Nobbe records an observation by 



