234 ASRICFtTURAL ESSAYS. 



rocks and other obstacles. Trials should be made of the advan- 

 tages which might result from ploughing flat land into ridges ; 

 and whether'ridge ploughing will not have a tendency to secure 

 grain from destruction by winter frosts. Attempts should also 

 be more extensively made, to raise winter wheat, which is 

 the most valuable of all grain ; also to obviate by culture, if 

 possible, the causes of the failure of wheat crops on our flat, 

 deep, rich soils, on which it is well known, attempts to ra se 

 wheat, successfully, have hitherto proved abortive, and 

 generally been abandoned. To discover also the best steeps 

 for grain and other seeds, to quicken vegetation and secure 

 them against smut and insects. What also may be the best 

 quantities of seed for sowing in different soils ; and the pecu- 

 liar advantages which might result from sowing different seeds 

 with a drill. Also whether dramed swamps are not the most 

 profitable lands, or what crops can be best raised upon them : 

 how lime as a manure will answer in our hot summers, and on 

 what kind of soil it is most profitable. 



In making experiments, care should be taken that we do not 

 draw conclusions too hastily. We ought seldom to do it from 

 a single trial. For a certain practice may answer well at one 

 time, owing to the peculiarity of the season, or some unknown 

 cause, which would not have the same operation at another 

 time. Too much ^confidence in smgle experiments, might 

 embarrass ormislead, rather than increase useful agricultural 

 science. 



If experiments are intended to make improvements, they 

 should be carefully recoided. For want of such records, much 

 useful knowledge is continually lost. Though many individu- 

 als have derived advantages to themselves from experiments, 

 but few have recorded them. Even those who make experi- 

 ments are liable to forget them, so as to give incorrect repre- 

 sentations of them when they attempt to relate them. 



Many useful discoveries therefore often die with those who 

 make them. To prevent these evils, either voluntary opera^ 



