Methods in Germination Tests of Coniferous Tree Seeds 79 



any given quantity of seed it is impossible to select an* average 

 sample which will be exactly representative, and it is further 

 impossible in practical work to have the conditions of germina 

 tion. such as moisture, light and temperature, exactly the same 

 for more than one sample. Furthermore, germination tests are 

 not an exact measure of the results to be expected in the field 01 

 nursery, but merely a relative index of the value of the seed for 

 this purpose. Germination tests in the greenhouse or laboratory 

 are carried on under the most favorable conditions, whereas, 

 conditions in the field are often highly unfavorable and very 

 rarely approach conditions under which the greenhouse germina- 

 tion tests are made. For instance, sudden changes in tempera- 

 ture, late frosts, drying out of the soil, "damping off" before 

 the seedlings appear above the soil surface are factors which may 

 be encountered in the field, any of which is sufficient to greatly 

 lower germination. Haack* found in working with pine seed 



*Haack: Keimung und Bewertung des Kiefernsamens. Zeitschrift fur 

 Forst und Jagdwesen, 38:465. 



that two seed samples which gave a 95 per cent final germina- 

 tion and a 65 per cent final germination in the laboratory yielded 

 but 39 per cent and 3 per cent respectively under unfavorable 

 field conditions. Furthermore, Haack's** experiments show that 



**Recknagel: The Equipment and Operation of a Prussian Seed Extract- 

 ing Establishment. Forestry Quarterly, 10:234. 



the number of plants obtained in the field is invariably lower than 

 the germination per cent. The period required for germination 

 will also usually be much shorter in the laboratory than in the 

 field. However, seed with the higher final germination and 

 especially the higher practical germination under laboratory con- 

 ditions can be relied upon to show the same advantages over 

 seed possessing these characteristics in a lesser degree when both 

 are sown under identical conditions in the open. 



Germination tests can be conducted at any time of the year 

 presupposing that favorable conditions for germination are pro- 

 vided for. 



METHODS. 



Seed tests may be broadly divided into two classes ; namely, 

 provisional tests and absolute germination tests. 



Provisional tests are merely rough estimates of the final 

 germination of any quantity of seed, made when it is impossible 



