CONTRIBUTIONS FROM THE WISLEY LABORATORY. 



237 



other half, which had carried no crop during the preceding year, by a 

 broad path. The fallowed land had been kept constantly free of weeds 

 by continual stirring of the surface soil, and had in 1906 carried a crop of 

 kale. 



Each half was divided into twelve equal plots, each measuring 36 feet 

 by 15 feet and having an area of 2 square rods. The plots were 

 separated from one another by paths 1 foot in width. The plots opposite 

 one another on each side of the central walk received similar treatment 

 (see plan, fig. 37). On Plots L, II., XIII., XIV. the soil was inoculated 

 with bacteria in the manner recommended (see below). Plots III., 

 IV., XV., XVI. were not treated in any way. Plots V., VI., XVII., 

 XVIII. received a light dressing of dung at the rate of 10 tons to the 

 acre, in order to test the question whether a slightly increased supply 

 of organic nitrogen would affect the work of the bacteria. Plots VII., 

 VIII. , XIX., XX. received a dressing of fresh slaked lime at the rate of 

 half a bushel to the square rod, in order that it might be ascertained 

 whether, if the inoculation on the untreated plots failed to produce any 

 effect, the failure was due to lack of sufficient lime, as it was known that 

 lime is present in very small quantity in the soil. In order to test the 

 question whether potash and phosphates were deficient, Plots IX., X., 

 XXL, and XXII. received a dressing of superphosphate at the rate of 

 2 cwts. to the acre, kainit at the rate of 2 cwts. to the acre, and lime as 

 on the plots last mentioned. Plots XL, XII., XXIIL, and XXIV. received 

 a dressing of calcium cyanamide (which contains lime in a caustic form 

 combined with nitrogen, and so supplies both lime and nitrogen) at the 

 rate of 2 cwts. to the acre, the amount being about equal so far as the 

 nitrogen is concerned to the addition of sulphate of ammonia at the 

 same rate. 



Culinary peas were chosen for the experiment, since they probably 

 form the most important leguminous crop in kitchen and market gardens, 

 and the produce is capable of being measured with considerable accuracy ; 

 it can be weighed, and one need not therefore depend upon personal 

 estimates of differences in the crop of the different plots, a method of 

 estimation which is always unsatisfactory. 



Four rows of peas were sown on each plot, the rows running north 

 and south. Each row measured 15 feet, and there was therefore on each 

 plot 60 feet of peas. Half the peas were inoculated with the bacterial 

 culture (see below) before they were sown and half were not inoculated, 

 the inoculated seed being sown on one plot and the uninoculated seed 

 on the adjoining plot that received similar treatment. There was 

 hus in the aggregate 360 feet of inoculated peas and 360 feet of 

 uninoculated on the well-cultivated ground, and the same amount on that 

 which had been fallowed, the total lengths of the rows of peas in the trial 

 being over a quarter of a mile. 



The Soil of the Experimental Plot.— The soil of the Wisley Garden 

 is well described by the Vegetable Committee of the Society, in terms 

 which will be appreciated by all who cultivate garden soil. The soil is 

 a "naturally poor, hungry soil, and requires constant manuring." It 

 overlies the Bagshot sand and is of a very sandy nature, poor in organic 

 mabter and containing a very small amount of lime. Indeed, a garden 



