XIII.] CONDUCT OF EXPERIMENTAL PLOTS 369 



ploughings, otherwise the manured soil will gradually 

 be displaced sideways ; if the plots are ploughed in 

 lands the furrows should be alternately gathered to and 

 cast away from the middle of the plot, or the manured 

 soil will be accumulated towards the middle. 



It is best to sow the manures (except the nitrates 

 and ammonium salts) a week or so before the seed and 

 plough them in. For sowing the manures, one of the 

 machines to be described later is best ; hand-sowing 

 produces considerable irregularities which can only be 

 obviated by mixing the manure with sufficient ashes or 

 burnt earth to make up a bulk of 10 or 12 cwts. per 

 acre and sowing the mixture in three successive opera- 

 tions. Calm weather must be chosen for sowing the 

 manures ; many fertilisers blow considerably if there is 

 the least wind stirring ; generally a few still hours may 

 be secured in the early morning. 



It will often be necessary, indeed always when the 

 manures are sown broadcast by hand, to have a screen on 

 the edge of the plot when sowing the strip which comes 

 up to this edge. At Rothamsted a canvas-covered 

 screen 16 feet long x 4 feet 3 inches high is carried 

 along the edge parallel with the machine or man sow- 

 ing. After sowing, the usual operations of cultivation 

 are carried out, but rather more care than usual should 

 be given to the singling of root crops, so as to obtain a 

 uniformly set out plant Gaps and misses can, on some 

 soils, be repaired by transplanting at this stage, but it 

 is not always desirable to do so, because one of the 

 properties of the manure under examination may be to 

 increase or diminish the tendency to lose plant In all 

 experiments with root crops the actual number of 

 plants on the plot should be counted before harvest 

 and recorded, as the figures often afford a means of 

 criticising the weight results, and of estimating the effect 



2 A 



