1922.] Trials of New Varieties of Cereals. 441 



from one race overtopping the other; or from lodging across the 

 strips; or from a more active root-system of one race than of 

 the other causing more vigorous growth of that race along 

 each line of division. 



It will generally be necessary to make sure of eliminating 

 tli is possible systematic error. This can be done by cutting out 

 before harvest one row of each race along the lines of separa- 

 tion. This will reduce the number of row s in each half-drill- 

 strip and involve either an addition to the length of the stup- 

 or a correction of the figures to give yield per acre. The 

 method of calculating the probable error of the experiment will 

 not be affected. The additional labour will be fairly consider- 

 able, but the work can be done well before harvest and will 

 facilitate cutting and harvesting. 



1 f the two races ripen so nearly together that they may be 

 cul on the same day, and if there is not much lodging, the 

 cutting can be done with a " side-delivery reaper " or with a 

 " self-binder " fitted with an adjustment to be referred to 

 later. 



If the difference in time of ripening is only a few da} r s, the 

 early ripening race may be left standing till the later is ready 

 for cutting, and in the case of barley this will generally be 

 possible. If, however, the difference is so great that there 

 would be any risk of " shattering " of the grain of the earlier 

 race, the strips must be cut by hand: also, obviously, if there 

 is severe lodging hand-cutting is the only feasible plan. Tf 

 the lodging is across the plots it may be necessary to go down 

 the spaces between the half-drill-widths with a stave and 

 throw back on either side the produce of each strip in order to 

 make a clean separation immediately before cutting. 



The writer's experience of the method has been only with 

 barley and although the straw was very heavy in both 1920 

 and 1921 in neither year was there any difficulty in cutting 

 with a " side-delivery " reaper. It is, no doubt, more likely to 

 be necessary to cut by hand in the case of oats than of either 

 wheat or barley. 



When the plots can be cut with a reaper or binder, if the 

 drilling has boon fairly straight, there will bo no difficulty in 

 driving the machine so as to cut each half-drill-strip separately. 

 This is facilitated if a man walks behind the machine and 

 warns the driver quickly if ho is going astray. The driver 

 should in any enso avoid cutting into the next half strip, and 

 if a few plants are left uncut these can be cut by hand and 



