127 



Chrysanthenvum, leucanthemum. — ^A vigorous plant in a grass field gave 

 about 26,000 seeds; six smaller plants at the same place gave from 1,300 to 

 4,000 seeds per plant. 



Sonchus arvensis. — Six plants in an oat-field were found to average 

 3,000 seeds per plant. On about four square metres 70 similar plants 

 were found. 



Matricaria inodora. — A specially vigorous, isolated specimen gave about 

 310,000 ripe seeds, which in six days germinated 97 per cent. The plant 

 gave thus about 300,000 germinable seeds. Another isolated plant gave 

 about 130,000 seeds. In later examinations the present writer has twice 

 found specimens giving each about 300,000 germinable seeds. 



Girsium arvense. — In a plant colony of this species, 25 powerful stems 

 "w^ere found per 0-4 of a squajre metre; some of these (the male plants) 

 gave no seeds, whereas the female plants on an average gave about 4,500 

 seeds per stem. 



In order to throw light upon how many weed seeds can he found in arable 

 soil, samples of soil were drawn in 1907 from fotxr fields in Jutland. 

 This was done by means of a four -sided iron frame 15 cm. high, so 

 that the samples comprised a layer of earth 15 cm. thick which corresponds 

 to the layer which in the selected district was subject to direct treatment 

 with plough and harrow. 



As the fields in question were supposed to contain many weed seeds, 

 one dare not take the resialts as an expression of how many weed seeds 

 Danish fields generally contain. 



The samples were washed out in sieves with meshes so small that all 

 weed seeds were retained. By an examination of carefully drawn average 

 samples of the content of the sieves, the species and the amount of weed 

 seeds in the samples of soil were determined. Seed of the following genera 

 and species occurred in greatest amount : — Chenopodium sp., Scleranthus 

 sp., Spergula sp.. Polygonum lapathifolium and Riunex acetosella. From 

 the results of the investigations, it was calculated that the fields contained 

 in the 15 cm. top layer of the soil 193,600, 116,600, 88,200 and 141,900 

 "weed seeds respectively per square metre. The seed of Chenopodium 

 found in one of the samples germinated 70 per cent., but the main part of 

 those species which were tested for germinating capacity germinated 

 between 20 and 30 per cent. Supposing the average germinating capacity 

 to be 26 per cent., and that the seed was evenly dispersed in the entire 

 layer of soil, the 2 • 5 cm. top layer in the four fields contained, according 

 to the calculation of the figures stated, the following numbers of germinable 

 "weed seeds per square metre : Nimaber 1 — 8,066; No. 2 — ^4,855; No. 3 — 

 3,674 and No. 4—5,913. 



V. — How do the weed seeds retain their germinating capacity in the soil ? 



The procedure in these experiments has, in all cases, been that small 

 flower-pots, in the middle of which ] 00 seeds mixed with soil were placed, 

 were buried at the beginning of the experiment. Each spring, a mmaber 

 of these pots were dug up and the seeds from them tested for germinating 

 capacity. The contents of the fiower-pots were spread in earthen bowls 

 which were, beforehand, almost filled with garden mould from a place where 

 the weed species under test had not grown in the memory of man. For 

 comparison 100 seeds of the same original sample — ^which was, in the 

 interval, stored dry in the rooms of the State Seed Testing Station — ^were 

 sown each year in a similar way as the above-mentioned. The main 

 results of the first series of experiments of this kind are to be found in 

 Table 3 {see page 135). 



In the experiments mentioned in this table, seeds were buried at 30 cm. 

 •only. As it is of considerable interest to see how the seed retains its ger- 

 jninating capacity at those depths of the soil which are generally cultivated. 



