PERSONAL EQUATION IN BREEDING EXPERIMENTS. 357 



TABLE IX. 



SHOWING TIII-: MEAN DEVIATION FROM THE MEA i. \\i> NET) AND TOTAL 



IATIONS OF THE COUNTS OF ALL OBSERVERS 





<>! ilu-i- (lilii-i'i-in c-s are very intcn -iin^. I'm- 

 No. I. had a hi^li "<' <l-\ i.iiimi on starchy krnu-U. u-ndin. 



iinatc tin- yellows and practically zero n i dr\i.itii>n "ii 

 -\\cri k. rnels; XI\*. is exactly the op . h.i\iir^ verj -in. ill 



net dt-xi.iiions on the starchy and tending -tnnnjv i<> over- 

 estimate the yellows among the sweet k(iiu-U No. II. li.id tlu- 

 iciidi-iKA to underestimate the yellow sweet>, and c< n-rc-ixuidin^ly 

 to ov< i' -lim.iic tin- white sweets. No. III. oni-i-M-niK umi 

 f-iini.iu-d ratht-r lu-a\ - ily all yellows and o\ --n -n'ni.iti-d all u hiic-. 

 No. IV. did ! I >' the opposite. No. \'. >hi\\ - a \-r\ irraiic 



se1 ! net di-ii-ilnitii>ns, o\vin^ to his idi<i-\ m IM-\ respecting the 

 di-t rimination i n starchiness and nin--t.in-hiii' Tlic 



n -ult i- that \\hilr he underestimated tin- vll<>\\ -tan h\ ki-nn-1-, 

 In- >\ -i\-- tiinat i-d all I he ol In -r rla-- No. \ I. -nicu hat ninlcr- 



estimated the yellow starchy, but \ n-c-tiniatrd tin- \ clln\\ sweet. 

 No. X shows an lAiraonlinarily Miiall net dc\ialion <ui tin- 

 >\\rd kcnii-l-. Inn di-tinctly unli-n>tiniatfd the \cllc\\ -larrhy. 

 In yrnt-ral the table shows in a striking way, that the indi\ iluality 

 i>t" tlu- i.li-i-i-\i-r i- . -v to IK- nrkiMu-d \\ith in work of thi- 



sort. 



