SEED CONTROL. 59 



germination of the former. Experiments conducted with a few 

 kinds of weed seeds tend to show that the seeds of some of our 

 native plants produce a small number of seedlings as compared 

 with those of cultivated plants. A large number of our worst 

 weeds are perennial and, although they produce seeds, do not 

 depend upon this method for self preservation ; hence their seeds 

 frequently show a small germinating capacity. A very large 

 proportion of our serious annual weeds, on the other hand, pro- 

 duce enormous numbers of seeds, apparently upon the principle 

 that, in some cases at least, but few of them will germinate. 

 Again, the conditions of germination of many weeds seem to be 

 different from those of cultivated plants, in that the seeds of the 

 former often require a long period of rest, and perhaps, also, 

 alternate freezing and thawing, before they will germinate. 



However, in such cases, while the injurious effects may not be 

 seen immediately, the weeds are likely to appear in future crops, 

 when least expected. Professor Goff, of Wisconsin, found that 

 seeds of " Redroot " (Amaranthus retroflexxis) would not ger- 

 minate at all when fresh, but after being kept for several years 

 in a bag in a dry place they sprouted quite well. 



On the other hand. Dr. J. C. Arthur, after conducting a test 

 with various weed seeds at the New York Experiment Station in 

 1887, came to the conclusion that some species became incapable 

 of germinating after being kept dry three or four mouths, and 

 therefore he thinks that the danger of fouling land from sowing 

 weed seeds with farm or garden seeds is not so great as commonly 

 supposed. Seeds of Russian thistle, collected October 30, were 

 tested in our laboratory the following May. After fourteen days 

 but ]o.o per cent had germinated. Seeds of the same species 

 one year and a half old failed to sprout at all during the same 

 period. 



The truth is that too little work has been done along this line 

 to warrant any definite conclusions. There is a field for study 

 heie which is well worth the attention of those engaged in Seed 

 Control. In general, however, it may be said that no seed should 

 be sold which contains as high as five per cent of weed seeds. 



In making reports on impurities, the European Seed Control 

 Stations take into account the kind of weed seeds present in a 

 sample. For example, in the Scandinavian Stations the follow- 

 ing, among otliers, are reckoned as " bad " weed seeds : cockle. 



