60 EXPERIMENTS UPON TURNIPS. [AppCndix, 



of the two manures may be cmisidered as practically equal — since this amount of 

 QilTerence may have arisen from the unlike qualities of the two plots of land, to 

 which the manures were respectively applied. 



This is an important practical rule for enabling us to judge accurately in 

 regard to the true effect of the several manures employed in the series of experi- 

 ments (p. 44) referred to, but the fact itself suggests also an important modifi- 

 cation in the mode of conducting all similar comparative experiments in future. 



In my previously published Suggestions, I have recommended the setting 

 apart o^ one undressed portion only of the field on which the trials were made — 

 considering that the produce of this portion would represent the average fertility 

 of the whole undressed part of the field But these experiments of Mr. Flem- 

 ing seem to shew that this opinion cannot safely be entertained. It appears to 

 be necessary, therefore, in all future experiments from which accurate deduc- 

 tions are intended to be drawn — that two undressed plots, at least, should, in earn 

 case, be measured out, and their relative produce ascertained, in order to afford a 

 trust-worthy average of the unaided fertility of the land. 



Suggestion I. — For the clearing up of this point, however, it would be very 

 desirable to institute a series of weighings of the produce of equal portions of 

 land, in several different parts of the same field, the whole of which has been 

 tilled and manured in the same way. This would throw some certain and sat- 

 isfactory light upon the amount of variation which, from natural causes, may 

 take place in the same crop, grown upon different parts of the same field, and 

 under the same circumstances. We should thus be enabled to allow for the 

 influence of natural causes upon the results of such experiments as are made, 

 with the view of determining the true action of the different manures we apply. 



Suggestion II. — But if some slight difference in the soil, which the eye cannot 

 detect, be capable of materially affecting the natural produce of the unmanured 



{)arts of a field, it may also be suflSicient to modify — that is, to increase or to 

 essen — the effects produced by the saline and other manures we apply to the 

 diflferent parts of the same field. It suggests itself, therefore, as the more pru- 

 dent and wary course of experiment to dress two plots at least with each of the 

 manures whose relative virtues we are desirous of testing, and these in different 

 parts of the piece of land upon which our trials are made. The mean produce 

 of the two or more plots we thus dress, compared with the mean produce of 

 those to which no dressing has been given, will indicate more nearly the aver- 

 age eflfect of the manure we have been tiying, upon the given soil and crop. 



The reader will perceive in the new precautions thus indicated, one of those 

 practical results which year by year will necessarily flow from the continuation 

 of the train of inductive experimental research, now, I hope, fairly entered upon 

 by the practical agriculture of our country. 



2°. Guano. — Among the other experiments upon turnips here stated, those 

 upon guano are the most practically successful. Thus, per acre, without any 

 farm-yard manure 



3 cwt. of guano alone gave 23 tons S cwi. oi Swedes . . (p- 44)- 



m blkl^fwooTles I ^^ '»- 2 =-■ or Early VeUo,. . (p. 44). 



5 cwt. of guano alone 32 tons 15 cwt. of White Globes . (p. 46). 



3i cwt. do. 20 tons cwt. of Yelhno t^ White mixed f p. 46}. 



3j cwt. do. 28 tons cwt. oi Purple-topped Yellow (p 47). 



These results are very gratifying, since they seem to shew that for the turnip 

 crop this light and portable manure may be substituted with safety for fann- 

 yard dung. But they are more gratifying in connection with the large reduction 

 which has lately taken place in the price of this substance. In norte of the 

 cases above mentioned did the quantity applied exceed 5 tons per acre. This 

 quantity may now be purchased for three guineas, though when these experi- 

 ments were undertaken it cost -CG 5s. 



It is no small matter of congratulation, tha. this important reduction has been 



