566 



Journal of Agricultural Research 



Vol. XIX, No. II 



Table XX. — Twenty lowest annual rankings and 3-year average rankings from 108 

 Early Six Weeks tuber lines <* 



I9I6 



Line No. Yield 



535 



545 



*6oi 



*539 

 *574 

 *536 

 *540 

 621 



559 

 538 

 555 

 573 



*542 

 *6o9 



*585 

 *6i9 



547 

 *546 

 *554 

 *5o5 



520 



Pounds. 



6, 969 

 6, 969 



7,550 

 7,550 

 8,131 

 8,711 

 8,711 

 8,711 

 8,711 

 9,292 

 9,292 

 9,292 

 9,292 

 9, 292 



9,292 

 9,292 

 9, 292 

 9,292 

 9,872 

 9,872 

 9,872 

 9,872 



3-year average. 



Line No. 



***62 I 



***545 



J553 



573 



***538 



*534 



**563 



*576 



**555 

 **566 



**559 



*52 7 



**520 



*574 

 *528 



*57o 



*535 

 *5i8 



**547 

 *599 



Yield. 



Pounds. 

 10, 027 

 10, 758 



12, 954 



13,39s 

 13,628 



13, 705 



13, 843 



14, 015 

 14, 324 

 14, 660 

 14, 750 



14, 828 

 15,041 

 15,331 

 15,461 



15,475 

 15,486 



15, 581 

 15, 598 



15, 796 

 15, 796 



a The number of stars indicates how many times the line has ranked among the lowest 20. Lines not 

 starred in columns 3 and 5 appear and are starred elsewhere in the table. Three appear three times, 8 

 appear twice, and 38 appear only once. 



It will be noted from Tables XIX and XX that in either case the 

 greater number of these lines have appeared only once in either of tliese 

 groups. line 561 appears in the low-yielding groups of 191 6 and 191 7 

 but is the highest-yielding line of 191 8. It also appears among the 40 

 lines selected in 191 8 as having the best vines. While the performance 

 record may not so indicate, it is undoubtedly one of the good lines of 

 this variety. Aside from this one line, none of those that have ranked 

 twice among the lowest 20 according to the 3-year average appear among 

 those selected in 191 8 as having good \-ine characteristics. We may 

 safely say that the perfonnance record is more reliable as a means of 

 eliminating the lowest-yielding lines than in pointing out the higher- 

 yielding ones. The following 191 8 field notes show that in this as well 

 as in the other varieties yield records will not eliminate lines with de- 

 generate tendencies. 



