Scientific Agriculture. 



48 



[January, 1910. 



Plot 12 over Plot 3 which is certainly 

 more than 8 but less than 12 per cent. 

 The mean error of a single year's re- 

 sult is, similarly 10 per cent. 



Table II. 



QQ 

 * <D 



CO 



98-1 88-8 



1906, 

 Mangolds 



1907. 



Wheat, 



1908. 

 Swedes. 



Mean of 

 5 years. 



95-8 



86-3 



92-8 



92-3 



90'6 



951 



94-9 



93-7 



99'2 



102-4 



100-2 



100-3 



105-0 



109-1 



114-9 



109-0 



109-2 



107-0 



97-3 



104-5 



o 



s 



A. .. 



B. .. 



C. .. 

 . D... 



E... 



Taking another example, Table II. 

 gives the results obtained in the last 

 five years on five plots in Little Hoos 

 Field which have received exactly the 

 same treatment ; in order to make it 

 easier to judge the figures, the actual 

 yields each year have been reduced to 

 a common standard, taking the mean 

 of all the five plots as 100. 



The experiment had to be started on 

 the assumption that all the plots were 

 exactly alike, and if so, the mean error 

 attaching to the result of a single plot 

 in any year is 7*5 per cent., but with 

 the five years' trials it is beginning to 

 be clear that there are some real per- 

 manent differences between the plots, 

 which improve from A. to E. Still, 

 whatever may be the real position of 

 each plot as revealed after further years 

 of experiment, we may expect in any 

 one year to find a particular plot 7 5 

 per cent, in error on one side or the 

 other. 



Space does not permit of the con- 

 sideration of more cases, but the general 

 result of the examination of many 

 series of experiments indicates that the 

 mean error attached to the yield of a 

 single plot is about 10 per cent, plus 

 or minus. In other words, if we have 

 three experimental plots giving yields 

 of 91, 100, and 110 respectively in any 

 one year — for example, 18, 20, and 22 

 tons of roots — it is not right to con- 

 clude that such differences have been 

 brought about by the treatment ; the 

 three plots must be considered as giving 

 equal results. Of course this figure 

 is obtained from a consideration of 

 the Rothamsted results only, and 

 other soils might be found on which 

 the conditions were so much more uni- 

 form that the experimental error will 

 be reduced and a closer agreement be- 

 tween duplicates would prevail. The 

 examination I have made of other data, 

 however, though they do not permit 

 of working out the mean error over 

 such long periods, yet lead me to sup- 



pose that a 10 per cent, error is near 

 the truth generally, and may be taken 

 as a safe guide for working purposes. 

 In the records of experiments a good 

 deal of strained arguments is often 

 spent in explaining results or drawing 

 conclusions from them when the dif- 

 ferences are much less than the 10 per 

 cent, which we have thus found to be 

 the average error attaching to a result 

 obtained under favourable conditions. 

 Much of this might have been spared 

 if the experimenter had kept clearly 

 before him the fact that nothing less 

 than 20 per cent, differences have much 

 significance in a single experiment. The 

 only way of reducing the experimental 

 error and obtaining a closer result is 

 to multiply the experiments, either by 

 repeating them year after year or by 

 increasing the number of plots, prefer- 

 ably both, because there may be constant 

 differences in the soil, while the season 

 also may induce variations in the effect 

 of the treatment. The first step, how- 

 ever, is to multiply the number of plots 

 set aside for each kind of treatment; 

 taking five plots irregularly distributed 

 about the field, we shall obtain in a 

 single year a result that is as accurate 

 as need be, except for special variations 

 induced by the character of the season. 

 Of course this means a considerable 

 increase in the amount of work attached 

 to the experiment. For example, instead 

 of six plots each of half an acre, we 

 ought to take thirty plots of a tenth of 

 an acre, six different kinds of plots and 

 five of each kind ; every plot would also 

 need to be harvested and recorded separ- 

 ately. Such a form of experiment is 

 necessary if small differences are to 

 be brought out, as, for example, the 

 differences that exist between the 

 various kinds of barley usually grown in 

 this country. 



Increased accuracy is not to be 

 obtained by increasing the size of the 

 plots ; it is questionable whether irregu- 

 larities of soil are likely to be more or 

 less pronounced on large plots, and with 

 very large plots one new source of error 

 is always introduced — the difficulty of 

 getting the cultivations, sowing, harvest- 

 ing, &c, of all the plots carried through 

 in the same day. As long as the plots 

 are above 1-40 acre size does not matter 

 much, and the size that is most con- 

 venient for the handling of the crop, its 

 weighing, storage, threshing, &c, should 

 be selected, always remembering that 

 it is by the number of plots only that 

 the error can be reduced. 



It is altogether wrong to take large 

 plots and then select small areas within 

 the plot for weighing. Such a proceed- 



