May 11, 1900.] 



SCIENCE, 



ni 



running through a plot will so change the 

 capacity of the area as either to vitiate the 

 results or render them exceedinglj^ unsatis- 

 factory. To avoid this and prevent the 

 mixing of treated soil the writer has resorted 

 to the spade. 



He has experienced great difficulty in 

 manuring the plots evenly. This is a prac- 

 tical, difficulty, not appreciated by the aver- 

 age workman, and if the dung is drawn 

 directly from the barnyard there may be a 

 great variation according to the amount 

 and character of the bedding used. The 

 manure should be made by only one class 

 of animals, as cattle or horses, or mixed 

 with the greatest care. Instead of being 

 spread upon the plots in winter it is better 

 to place it in piles and when well rotted 

 and forked add by weight or measure, pre- 

 ferably the latter, to the plots. With com- 

 mercial fertilizers the above difficulty is 

 eliminated. 



During last winter a test was made of 

 placing the piles of manure upon some of 

 the plots for the purpose of determining the 

 influence of the same upon the soil and crops . 

 In every case the growth was much retarded 

 and in some instances the seed did not pro- 

 duce more than a small fraction of the full 

 quota of plants. The results of an experi- 

 ment would have been worthless for any 

 plot where a manure heap had been upon a 

 portion of the ground. ' 



It is not the purpose here to discuss the 

 cause of the failure where the eni'iching 

 was excessive, but the fact remains and any 

 one who would follow plot experimentation 

 must not fail to manure his land evenly 

 with the greatest care and prohibit the piling 

 of dung upon the plots. 



Previous Crops. — One could have all the 

 points in common save that of the previous 

 crop and the results might be worthless. 

 This the writer knows to his sorrow. The 

 influence of a crop is more lasting than one 

 might suppose. Of course the wide difier- 



ence between sod ground and cultivated 

 land goes without saying ; but let both pre- 

 vious crops be tilled ones, as for example, 

 beans and corn, or potatoes and onions and 

 there is enough of the personal equation to 

 make the land unsuited for plot experi- 

 ments unless the plots are confined to some 

 one crop. When a crop is successive no 

 portion of that experiment should grow 

 upon old land. The marked influence of 

 any leguminous crop is a case in hand. 

 The soil from an old pea, bean or clover 

 field when added even in small quantities 

 to land new to such crops may double the 

 yield. Turnips after turnips may be so 

 diseased as to neutralize results of cultural 

 treatment when compared with the same 

 crop upon new land. 



All prospective ground for a plot experi- 

 ment ought to be uniformly in the same crop 

 or succession of crops for a term of at least 

 five years and fallow for the last season. 



That the quality of the seed has an influ- 

 ence upon the results goes without further 

 saying. lu the early days of the writer's 

 experiments the supply of seed would some- 

 times fail and that of another lot, but of the 

 same variety would be used to finish the 

 area. With bush beans, for example, it is 

 very difficult to get the same strain of a va- 

 riety from different dealers. The same is 

 true of potatoes and nearly all sorts of truck 

 crops. It is of considerable importance, 

 therefore to purchase so liberally that there 

 is no chance of one plot not receiving from 

 the same lot of seed as all the others. 



As it is our practice to grow two and even 

 three crops as of peas and beans, during a 

 season, it is imperative that enough seed be 

 purchased in the spring to supply the needs 

 of all the sowings and plantings of the grow- 

 ing season. 



In planting and sowing there is a possi- 

 ble element of variation. After the seed 

 bed has been made as uniform as possible 

 throughout the whole area, the sowing or 



