80 REMARKS UPON PRKCKDINa EXPERIMENTS. [Appendix^ 



each of these pairs of experiments, are so great that had they been ohtained from 

 plots of ground dressed with different manures we should readily have ascribed 

 them to the unlike action of the substances we had applied. Doubts may natu- 

 rally arise, therefore, when we look at the several tables of results contained in 

 this Appendix, how far the differences presented in them are really due to the un- 

 like action of the manures employed, and how far to natural causes not hitherto 

 investigated. Can all the experiments madedui-ing these last two years with so 

 much care really be vitiated oy this source of error'? The point must be eluci- 

 dated by further experiment. Should it prove that we have here a general 

 source of error, it is satisfactory at least that we have discovered it at the threshold 

 as it were of our accurate experimental inquiries, and that we can devise means 

 of avoiding it in future. 



I therefore repeat the Suggestions I. and II., which I ventured to offer in page 

 60 (Appendix), that some of my readers, of whom I believe many are interested 

 in this subject, would in the ensuing season ascertain accurately the produce of 

 equal measured quantities of the same field, under whatever crop it may be, 

 and publish or transmit the result to me — and that in all future experiments 

 made with the view of ascertaining the effect of different manures upon any crop. 

 two plots at least, and not adjoiiiing to each other, should be treated alike in each 

 field, and the mean of the several results obtained with each substance taken as 

 the average produce from which their comparative effects are to be estimated. 



These points appear to me to be of primary importance, and to lie at the 

 foundation of the structure 1 hope we are now beginning to rear with the results 

 of inductive experimental agriculture. 



4. Action of soot.— In these experiments a top-dressing of soot increased con- 

 siderably the produce of oats and wheatj while it diminished the produce of po- 

 tatoes when mixed with the manure. Thus the produce per acre on the dressed 

 and undressed parts was — 



Oats. Wheat. Potatoes. 



Undressed . . 49 bush. . . 44 bush. . . 11 tons 10 cwt. 

 Dressed ... 55 bush, . . 54 bush. . . 11 tons 3 cwt. 

 The unfavourekble effect upon the potato crop may probably be due to the 

 mode in which it was applied, as in other districts it is veiy useful to potatoes, 

 and gave, as we have seen, when applied alone to turnips, an increase of 4 tons 

 per acre. (See Mr. Fleming's Experiments, Appendix, p. 43: also, Lecture 

 XVII. p. 438). 



5. Comparative action of soot and of nitrate of soda. — The immediate effect 

 of both these substances is to darken the colour and to increase the growth 

 of hay and straw. In this respect the advantage is rather on the side of the ni- 

 trate, while the soot in some cases gives a little more grain. Thus the increase 

 of produce per imperial acre of the three crops of hay, wheat, and oats, dressed 

 ■with each of the three, was nearly as follows: — 



Hay. Wlieat. Oats. 



Grain. §traw. 



Soot .... 7 cwt. . . 10 bush. . . 6 bush. 6 cwt. 

 Nitrate of soda . 9 cwt. . . 10^ bush. . . 6^ bush. 7 cwt. 

 In both cases, however, the sooted grass was lighter per bushel. Thus their 

 comparative weights were — 



Wheat. Oats. 



Sooted . . , 58 lbs. . . 41 lbs. 

 Nitrated ... 62 lbs. . . 42^ lbs. 

 Nevertheless, the advantage to the practical man is decidedly on the side of 

 the soot, since the cost of 40 bushels of soot per acre was only I2s., while that 

 of 1 cwt. of nitrate of soda was 25s. It is only to be regretted that soot is so 

 variable in its constitution that firm reliance cannot be placed upon the uniform- 

 ity of its effects, 



6°. Actixyn of guano. — In the text, p. 460, 1 have stated the apparent conclusion 

 to which the Erskine experiments, taken in connection with all the others Ihav« 



