DECEMBER. 5Q5 



cultivation, and attended with as much loss in head- 

 Jands and borders, as they are ruinous to any exact- 

 ness of accompt. 



But as many persons keep accompts without at* 

 tending to this point, I would observe, than when 

 all the wheat, all the barley, all the oats, &c. are re- 

 spectively thrown together, some very essential ob- 

 jects of experience depend on guesses, which ought 

 to be ascertained correctly. Has fallow, or clover, 

 or beans, paid best, as preparations for wheat ? 

 How is that question to be answered, if all are hud- 

 dled together in one barn or stack, and meet in the 

 same accompt ? The farmer can guess nearly. He 

 may : but go to a chemist, ask him if his science 

 was pushed to the present perfection by accepting 

 such guesses, instead of experiment ? besides, they 

 are in their nature quite uncertain ; and when a 

 comparison is formed by two guesses, a very little 

 error in each will amount to so much in both, as 

 to overturn all authority. Another point is, a man's 

 guess being influenced by a favourite theory : a 

 rigid friend to fallows, when he draws, by guess, a 

 comparison between them and beans, will be apt, in 

 the nature of things, to be partial : he should not 

 put himself in the situation : he who would abhor 

 the idea of falsifying a fact that is before him, might 

 guess, at least, without sufficient accuracy. 



If the fields be not very small, the inconvenience 



of keeping crops separate is little. Stacking corn is 



better understood and executed in the Isle of Wight 



4han in most other parts of the kingdom : a great 



Q q 2 stack 



