165 



When the work with grasses began at Svalof the first efforts were 

 directed toward collecting for investigation a large number of individual 

 plants which possessed sufficient hardiness. This was not difficult since all 

 the grasses under consideration grow wild in Sweden. Promising individuals 

 were found in pastures, meadows and waste places while others were obtained 

 from cultures produced from commercial samples. In this way several 

 thousand individuals have been brought together during the past few years 

 although large numbers which proved less valuable have been cast aside. 



Photo, by L. H. N. 



FIG. LIII. Sowing Orchard Grass Multiplication plot in drills 0-50 m. apart 

 (Svalof, Aug. 10, 1910.) 



Experience in the study of grasses thus far seems to indicate that the possi- 1 

 bilities of effecting improvements of practical value seem quite as great as, 

 if indeed not greater than, in the case of cereals. 



Many types of grasses have been distinguished which possess outstanding 

 morphological characters which are of value as a means of identification. 

 Thus the different degrees of hairiness of glumes, peculiarities in the types 

 of panicle in Orchard Grass and degrees of awn development on the 

 flowers of Timothy are all useful for the above purpose, although every 

 imaginable combination and gradation may be found. 



