148 TESTING OF SEEDS. 



Thus the average number of days for each seed is 7^ and, the 

 percentage germinating heing 70, the quality of the seed may he 



70 

 expressed by the fraction y' 



The number of seeds tested must, of course, be sufficiently large 

 to guarantee the trustworthiness of iiie results, and it is obviously 

 most convenient to experiment with 100 seeds or some multiple of 

 100. It is unnecessary to wait until the entire process of germina- 

 tion is over in order to consider the test complete ; the germinative 

 power of the seed is sufHciently proved if the germ has swelled up, 

 the plumule and radicle have already attained a certain develop- 

 ment, and the cotyledons, and the perispenn, if any, are perfectly 

 healthy. The water used should, if possible, be soft, such as rain 

 or snow or distilled water, or it may be taken from a small running 

 stream that does not flow over limestone rocks. It is evident that 

 the progress of any experiments undertaken should be watched from 

 day to day. 



In favourable weather the seeds may be sown out in the garden 

 in ordinary soil ; or, if they are small and difficult to find again, 

 they may be enclosed in a fold of flannel or any absorbent cloth 

 before being put into the soil, so that they may be easily examined 

 from time to time and the progress of germination constantly watch- 

 ed. Enclosing the seeds in the flannel or other absorbent cloth has 

 also some other important advantages : the seeds are less Hkely to 

 rot from excessive damp, or to suffer from insufficient aeration or 

 from rapid alternations of heat and cold, or be attacked by insects, 

 than if they were in direct contact with the soiL 



An improvement on the preceding method is to sow in ordinary 

 garden pots, which, being portable, can be moved into the sun or 

 shade or under shelter according to necessity, so as always to secure 

 for the seeds the best conditions for germination. Moreover, sown 

 in pots, the seeds are much less liable to be injured by excessive 

 damp. Some special precautions have to bo employed in the use 

 of pots. After placing a potsherd over the hole at the bottom, the 

 pot should be half-filled with coarse sand or potsherds or tiles 

 broken up srnall, in order that no water may collect and stagnate 

 at the bottom of the pot. Good garden soil should than be put in 

 and the seeds sown in it to the proper depth, which will vary 

 directly with the size of the seed. Instead of watering from above 

 and causing an impermeable crust of fine mud to form on the sur- 

 face of the soil, the pot may, whenever necessary, be placed for a 

 sufficient time inside another larger pot containing water. Tlw 



