38 FORAGE CROPS IN DENMARK 



Although other characters of the roots were noted, the classifica- 

 tion of the samples was based exclusively on the quantity of 

 food-stuff (total solids) per acre produced. When the roots had 

 been weighed and analysed and the number of cwts. of solids 

 per acre calculated for each sample, the difference between the 

 highest and the lowest yield was divided in three equal parts. 

 Those samples that fell in the highest third were considered of 

 Class I., those on the lowest third were Class III., the middle 

 being Class II. This being done at each station the results were 

 tabled as shown on the opposite page, the samples being placed 

 in the order of the yield per acre calculated for each sample 

 as the average of the yields found at the three stations. 



It will be seen that the three best samples were placed in 

 Class I. at each station, and that the two poorest ones were found 

 to be in Class III. at each station, but that the positions of the 

 others varied somewhat from station to station. This is partly 

 unavoidable, as the different soil and climate at the various 

 stations affected differently those samples which were on the 

 border line between the classes. But it is also * partly due to 

 shortcomings in the technique, which was therefore gradually 

 improved. 



The present system of working is as follows : The various 

 kinds of roots examined are distributed between the seven 

 Experimental Stations and three branch stations in such a way 

 that each kind is grown on at least five stations. At each 

 station each sample is sown on at least twelve plots spread over 

 the field, each plot being at least ten square yards. Each plot 

 consists of two rows. The distance between the rows is 60 cm. 

 (24 inches), the length of the rows may be from 10 to 12 m. 

 (11 to 13 yards). The roots grown on the outside borders of 

 the field are not used for the experiment, but are used as a kind 

 of screen. When two strains of different growth, for instance, 

 with small and large top, are in rows next to each other a screen- 

 ing row of each kind is interposed. These rows are not included 

 in the experiment, but simply grown to make sure that the two 

 rows of the strains forming part of the experiment are grown 

 under normal conditions. The sketch on p. 40 of part of the 

 field at Askov Experimental Station will show the arrangement. 

 The field is divided into four quarters ; each quarter is divided 



