90 EXPERIMENT STATION. [Jan. 



This table shows, iu the case of both varieties, fairly large co- 

 efficients, and they are in both eases larger for the short vines 

 than for the long vines. These ligiires form a too slender basis 

 for a definite conclusion as to the correlation between vine 

 length and the average weight of peas produced, but so far as 

 they go they consistently favor the supposition that the correla- 

 tion does exist and is fairly large. The average of the four is 

 +.2287. 



Table VI. — Correlation of Average Seal Weight and Vitics produced. 



Strain A, D, F, G, K, + .0710 ± .0234 



Strain B, E, II, J, + .1045 ±z .0286 



Alaska, + .0140 =t .0178 



First of All, — .0290 zt .0390 



Turning now to the consideration of the question as to 

 whether the heavier peas produce larger vines than do the lighter 

 ones, we have the figures shown in Table VI. For reasons 

 shown a little further on in this paper, the group Excelsior I. is 

 divided into two strains, one of 5 lines and the other of 4 lines 

 as shown in the table. The group First of All is from a selec- 

 tion of this variety that is of the same nature as the others. The 

 number of vines is relatively small, and the figures, therefore, 

 of less value than the other groups. It is because of the snuill 

 numbers that the correlation of parent vine length and weight 

 of their seeds are not given, but as far as they go they are in 

 reasonable agreement with those of the two groups that arc 

 given. 



It appears from the limited data given in Table VT. that the 

 correlation is larger for the wrinkled Excelsior peas than for 

 the starchy Alaska and First of All varieties. Only the coeffi- 

 cients for the first two groups should therefore be compared 

 with the correlation of about +.06 found to exist between parent 

 and offspring as shown in Table II. 



No positive conclusion in this matter can be drawn. The in- 

 dications are that a part and possibly all of the correlation of 

 .00 may be accounted for by the correlation between length of 

 vine and seed weight. 



