WEED SEEDS IN THE SOIL. 377 



to hold their own, however, especially the first two, in the open fields where 

 competition is more severe. 



The difference in the number of viable seeds found on well tended ground 

 (Purdue plot) and neglected land is very pronounced, showing clearly the 

 value of frequent and thorough cultivation in the control of weeds. Better 

 cultivation of the fields and regular mowing of weeds on fallow and waste 

 land would considerably reduce the ten million dollar loss, which is occasioned 

 yearly in Indiana by the numerous weed pests. 



The permanent sod contained, as shown in the table, a comparatively 

 small number of viable weed seeds per square foot. 



The second table shows the species and the number of weed seeds con- 

 tained in the separate inch layers of sample number 6. The figures indicate 

 that the largest number of weed seeds occurs in the surface layer. A large 

 majority of the total number of seeds are contained in the first six inches 

 of the soil layer. This latter fact indicates the depth at which the land in 

 question was usually plowed. Beyond this depth the number of viable seeds 

 is considerably reduced, leaving only four seeds in the twelfth inch, in this 

 particular case, as compared with fifty-two in the sixth inch. At a greater 

 depth than twelve inches the number of weed seeds, found on upland soil, 

 is doubtless very small. On overflow land, however, it is quite likely that 

 many viable seeds are found at considerable depths. 



The third table summarizes the total number of weed seeds, calculated 

 per acre, as reported by the investigators mentioned in this paper. 



A brief reference to this work has been made in the twenty-eighth Annual 

 Report of the Purdue University Agricultural Experiment Station, pp. 31-32. 



