WixiFUKD E. Brexciilky 20 



cultivated out and prevented from seeding. Also, as the field is experi- 

 mental, an extra amount of weeding is done during the growing seasons, 

 and this also tends to reduce the quantity of weeds. Added to this, the 

 soil is now very poor owing to the long continued cropping without 

 manure, and this impoverishment influences the growth of weeds as well 

 as of crops and tends to reduce their vigour and abundance. All these 

 limiting factors are reflected in the number of buried weed seeds which 

 germinated in the soil samples. 



From the two holes cos-ering i sq. foot only 43 arable weeds were iden- 

 tified, though probably some of the 39 seedHngs which died in infancy 

 without developing belonged to the same category. These numbers are in 

 sharp contrast to those obtained from Long Hoos where no experimental 

 interference took place and where the land received manure at intervals 

 in the ordinary course of management. 



During the first few months to Dec. I'JlG, Alriplcx paluhi and Chcno- 

 podium album were the only weed seeds that germinated, and few even 

 of these were present. In the spring of 1917 the number of these species 

 increased considerably, and various other plants came up, bringing the 

 total arable species to eight. Several plants of Arenaria scrpi/lUfolia and 

 Matricaria inodora appeared, but the other species only occurred in ones 

 or twos. Trifoliuiu jmilense was present in far greater quantity than the 

 weed seeds, but this is probably the result of an accidental error, as a 

 few years ago clover seed was sown on the fallow half instead of on the 

 clover half of the field, and it has not yet been possible to entirely elimi-' 

 nate the sj^ecies. Both Trifoliuin pratense and T. re pens were present in 

 quantity from six to eight inches below the surface, and consequently it 

 is probable that many of the seedlings were derived from seeds that had 

 been buried for some considerable time and had been worked down by 

 ploughing and cultivation. 



General .Discussion of the Results. 



A survey of Tables I-X shows at a glance how closely the flora de- 

 rived from buried seeds is associated with the history of the land. Per- 

 manent grassland is practically devoid of arable weeds, and also the 

 number and varietv of the species that do occur is greatly influenced by 

 the fact of grazing or cutting, as the case may be. Continual close grazing, 

 as on the Common, hinders seed production, and so reduces the number of 

 viable seeds that become buried in the soil. Continual mowing, as on the 



