212 PROCEEDINGS OF THE CHEMICAL DEPARTMENT. 



different soils and crops, and resting content with this informa- 

 tion, which, of course, will hold good only for the individual 

 substances used ; or by endeavouring to ascertain the mode in 

 which they produce their characteristic actions. In the former 

 case the results are empirical, but they are arrived at with com- 

 parative rapidity, so that it is possible, after several years' trial, 

 to say whether, for example, guano, bones, or superphosphate 

 produce the best effect on such and such soils, but they leave us 

 completely in the dark as to how or why they act as they are 

 found to do, or of the effect likely to be produced under circum- 

 stances different from those under which the experiments were 

 made. In the latter case the manure is not looked at as a whole 

 only, but we endeavour to trace the mode in which it acts on the 

 oil, and the effect due to each of its constituents, and thus to 

 make, if we may so express it, an analysis of its manurial effects. 

 Experiments conducted in this way for a sufficient length of time 

 to admit of satisfactory generalisations, enable us to decide what 

 part of the total effect of a manure is due to each of its consti- 

 tuents, and to foresee with certainty the result which would be 

 produced if their proportions were altered. 



The great difference between these two modes of conducting 

 experiments lies in the fact, that the first, even if conducted only 

 for a single year, offer some kind of conclusions, although they 

 may not be very certain, and may sometimes mislead, while the 

 last require to be continued for some time before any inferences 

 can be drawn from them, and they rarely become valuable until 

 the results of many successive seasons have been accumulated, 

 and compared with one another. As a natural consequence, the 

 former are much more attractive to the experimenter than the 

 latter, because he sees at once to what they lead, and has the 

 pleasant satisfaction of drawing some conclusions from them. 

 The other kind of experiments can only be prosecuted with success 

 by those who profess a real enthusiasm, and who are determined 

 to accumulate the results of many years, during which facts are 

 laboriously accumulated, from which no conclusions can or ought 

 to be drawn until they are sufficiently numerous. The experi- 

 menter must therefore exercise much self-denial, and be content 

 to wait patiently for many years ; to obtain results which he 

 sometimes cannot understand, or which appear to be irreconcile- 

 able ; to encounter failures of the most disappointing kind, 

 dependent on causes over which he has no control ; and perhaps, 

 after many years devoted to laborious inquiries, he may find that, 

 if he wishes to obtain absolutely'' cou elusive results, he must 

 commence a new series of experiments, so arranged as to clear 

 up the difficulties of those he has already made. 



In general the necessity for carrying on experiments for a 

 succession of years is sufficient to deter many from undertaking 



