238 



JOURNAL OF THE ROYAL HORTICULTURAL SOCIETY. 



soil more suitable for the experiment, so far as its natural poverty goes, 

 would have been difficult to find. 



The site selected was as nearly level as could be obtained, but had 

 a slight slope from south to north. The depth ol soil in the two halves 

 naturally differed considerably. The soil on the part that had received 

 ordinary garden cultivation (the eastern half) had been deeply trenched 

 for celery in 1907, and was therefore much deeper and much better worked 







3f& 



S*6 



7'* 





















u 





























2s 





























IL 















22 















is 















I 















s 



3 















b 



Si 



s 







fjfi8 







23 f 



I'h.. 88. -DlAflRAM SHOWING THE YIELD ON EACH OF THE PAIRS OK PLOTS. 



ili.ni the fallowed ground, which had been hoed only through 1907. Both, 

 of course, were dug alike before the sowing of the seed in the spring. 

 One point was particularly noticeable on the fallowed ground: the soil 

 there was appreciably deeper at the northern end, the bottom of the slope, 

 than at the southern end, owing to the washing down of the finer particles 

 of Boil daring the heavy autumn and winter rains of 1907-8, and this had 

 a marked effect up n the yield in the fallowed land. The diagram (fig. 38) 

 will show graphically what this effect was 



