44 EXTENT OF AN EXPERIMENTAL GARDEN. 



its being different every year, may better fulfil all the conditions 

 which appear necessary to the utility and accuracy of the results. 



But, in regard to a purely experimental piece of ground, two 

 considerations tend to throw some light upon the extent in 

 acres which it ought to include. 



1. Land experimented upon ought to be in a uniform, natu- 

 ral, and well understood condition. That is to say, its agricul- 

 tural history and past treatment should be well understood, and 

 the portions reserved for after experiments should be kept in a 

 similarly well understood condition. 



2. A second experiment must not be made on the same spot, 

 except as a sequel or continuation of a former one, until several 

 years have elapsed. By such a lapse of time, all appreciable 

 action of the first application may in most cases be expected to 

 be removed from the soil. 



Now, suppose thirty-two field experiments, eight of each of 

 four crops, to be conducted yearly over as many quarter-acres of 

 land, this would require eight acres in arable culture ; and, sup- 

 pose five years to be necessary in order that the influence of 

 the first application may die away, this period of time may 

 elapse before a new experiment could be commenced on any 

 one of the above plots. Thus five times eight, or forty, acres of 

 a given quality of land or eight acres of each of five different 

 qualities would be the full extent of arable land which such a 

 farm or garden would require. A considerable proportion of 

 this land would probably always be in grass, yet ten or twenty 

 acres additional of old grass-land would be necessary for ex- 

 periments to be made either upon itself or upon cattle or sheep 

 to be in part supported by it. 



No one, who considers how much thought and attention will be 

 required by each experiment, made with a view to truth alone, 

 and how numerous the secondary or subordinate experiments 

 which would accompany the principal ones and who adds to 

 these the experiments on the feeding and rearing of stock, and 

 the analytical researches which must accompany all the experi- 

 ments throughout the whole year will fail to see that from 

 forty to sixty acres of land so employed would give ample occu- 

 pation to any two skilful, industrious, and clear-headed men, 



