58 MASSACHUSETTS HORTICULTURAL SOCIETY. 



crops according to their predominating requirements, and adapting 

 formulas to the several classes. For general and indiscriminate 

 use, where the character of the soil and other special conditions 

 cannot be taken into account, this plan is open to less objections 

 than any other. 



But, even if our knowledge of the laws of plant nutrition were 

 sufficient for accurate calculations, as it is not, this plan would still 

 leave several very important factors out of account. The weightiest 

 is the supply which the soil can itself furnish. The next step in 

 advance, then, must be to test the soils. 



The faimers near Middletown (where Professor Atwater resides) , 

 need potash ; in other places, potash does little or no good. How 

 shall each of us ascertain what our soils most need ? There is no 

 better way than to trj' experiments on a small scale before making 

 large investments ; to put the question to the soil with different fer- 

 tilizers and get the answer in the crops. 



The best agricultural chemists in this country and in Europe, are 

 agreed in this opinion, Stoeckhardt, Knop, Woltf, Ville, and 

 Voelcker, have all recommended farmers to try experiments to test 

 the action of different fertilizing materials upon their soils. Of 

 course, there are soils so uneven that experiments on a small scale 

 cannot be decisive. Nor will a single crop tell the whole story. 

 To get the best results, the experiments should be carried through 

 a series of years,- — through a course of rotation, for instance. 

 The soil should be without previous accumulations of fertilizing 

 matters to interfere with the action of the fertihzers. By proper 

 care, trials may be made which will bring valuable results at com- 

 paratively little cost. 



Acting on these principles, the Connecticut Experiment Station, 

 of which Prof. Atwater was Director, in the spring of 1877, 

 sent out to as many farmers, nearl}' fifty sets of fertihzers, with 

 directions for experiment. It was recommended that these experi- 

 ments should be tried on long parallel strips of ground, as likel}' to 

 give greater average uniformit}' of soil than the same area in a 

 square. Nearly twenty of these farmers have reported, and their 

 reports are in accordance with what has been said. There is a great 

 difference in soils, and what produces the desired effect in one case, 

 may not in another. 



Prof. Atwater described an experiment made bj' D. H. Birdsey, of 

 Middlefield, Conn., on six plots, the first of which was manured 



