100 PRINCIPLES OF GARDENING. [CH. II. 



riments, far more forethought and care are requisite 

 than are usually bestowed upon them. 1. A space 

 should be left without any manure being applied, 

 othei'wise there will be no satisfactoiy basis of 

 comparison. 



2. The larger the space subjected to experiment 

 for each manure, the more entitled to confidence 

 vdW be the result. The reason for this is, ob^-iously, 

 that no two seeds will produce plants of precisely 

 equal prolificacy. Imperfect ripening of the parent 

 seed, yariance in the depth at which the seed is 

 buried, and many other circumstances, will be more 

 liable to have a controlling effect over the weight of 

 the produce from a small plot of crop than from a 

 larger. A dozen super-prolific, or defective plants, 

 on a square rod of gromid, v,i.\\ have an influence 

 on the result when calculated per acre, that would 

 be scarcely appreciated, if the experiment were 

 made on an eighth of an acre. 



3. If manures in solution are employed for soak- 

 iuo- the seed, a similar quantity of seed of the same 

 sample should be soaked for a similar length of time 

 in simple water. If liquid manures are given expe- 

 rimentally to plants during their growth, other 

 plants of hke number and growth, and in every 

 respect treated similarly, should at precisely the 

 same time have simple water applied to them. 



4. There should be a certainty tliat the manure 



