113 



chinch bugs in one plot and not in another, a sandy spot in one and not 

 in another, imperfect germination of seed in one plot causing a thin 

 stand, while in other plots the canes stand close together, and it will 

 be see i that the task of differentiating between varieties by growing 

 them in plots and submitting the canes produced to analysis is by no 

 means an easy one. It is a very complex problem. One season's work 

 should never be held conclusive ; a variety may have been placed at a 

 disadvantage from some one of numerous possible causes. 



In the work here the varieties were analyzed as often as possible, to 

 avoid the error of having analyses of cither unripe or overripe canes to 

 compare with the analyses of other varieties at their maximum; the high- 

 est analysis in the series may be taken as the basis of comparison. The 

 error of sampling was avoided as much as possible by taking good sized 

 samples, and by having them all taken by one and the same person.* 

 The errors arising from differences of growth were augmented, unfortu- 

 nately, by irregularities in the time of planting; some lots of seed being 

 received very late in the spring. The time of planting is noted with 

 each plot. 



EARLY VARIETIES. 



Several of these gave very satisfactory results, so far as early ripening 

 was concerned. The late date at which the laboratory was established 

 at the station did not admit of many analyses before they had passed 

 their maximum of maturity. In the case of the Early Tennessee and 

 Whiting's Early Variety this point had probably been passed before 

 any analysis was made. 



-Mr. Dentun did all the- sampling himself. 



1405G Bull. 20 8 



