952 
Journal of Agricultural Research 
Vol. XXIII, No. ia 
tubers of this lot were relatively long and narrow. The Early Ohio po¬ 
tatoes produced at this station since that time, that is in 1917, 1918, 
1919, and 1920, have been relatively short and wide. Lot 4 was obtained 
from Faribault, Minn. The tubers of this lot were elongated and were 
apparently somewhat variable in form (PI. 1, B). A few tended toward 
an oval form. Lot 6 was obtained from the region about Duluth. The 
tubers of this lot were relatively short, very broad, and considerably 
flattened (PI. 1, D). Lots 7 and 8 were obtained from growers in the Red 
River Valley. The tubers of these lots were characteristic of the more 
elongated cylindrical form of the Early Ohio of this region (PI. 1, C). 
Table II .—Showing that poor growth conditions at Crookston in igij as compared to 
Duluth and Grand Rapids did not affect the yield of Crookston stock in igi8 
I, lot 
No. 
II, source of seed, 
1916. 
Ill, yield. 
1917. 
IV, yield at 
University 
Farm, 1918 of 
seed grown in 
1917 at— 
V, yield at 
Duluth in 19x8 
of seed grown 
in 1917 at— 
VI, yield at 
Grand Rapids 
in 1918 of seed 
grown in 19x7 
at— 
VII, yield at 
Crookston in 
1918 of seed 
grown in 19x7 
at— 
Duluth. 
Grand Rapids. , 
Crookston. 
i 
Duluth. 
Grand Rapids. 
Crookston. 
Duluth. 
Grand Rapids. 
Crookston. 
Duluth. 
Grand Rapids. 
i 
i 
0 
i 
3 
Q 
Grand Rapids. 
Crookston. 
2 
Anoka. 
203 
326 
44 
293 
271 
202 
i 93 
138 
143 
290 
299 
269 
3 
Grand Rapids... 
174 
318 
63 
242 
224 
263 
220 
240 
4 
Faribault_... 
174 
311 
80 
288 
224 
315 
220 
XIO 
146 
6 
Duluth. 
219 
314 
75 
297 
293 
308 
165 
158 
169 
246 
247 
311 
30; 
249 
255 
7 
Glyndon.. 
202 
316 
7 i 
346 
227 
271 
152 
178 
139 
256 
266 
359 
273 
155 
299 
8 
Hawley. 
179 
265 
56 
264 
2X8 
288 
314 
255 
256 
258 
191 
235 
A part of each of these lots were grown at University Farm, Duluth, 
Grand Rapids, and Crookston in 1917. The form of the tubers produced 
at each of these places was similar in all the lots. It was evident that 
regardless of the form of the tubers in the original lots the growth condi¬ 
tions at each place produced a uniform distinctive form of tuber. The 
1917 crop was carefully studied for possible differences in form between 
the lots grown at any one place. It was thought that slight differences 
in form were observable between some of the lots grown at Duluth. 
These differences in form were not found to remain consistently in 1918, 
although a similar amount of variation between the lots was observable. 
In 1919 a study was made of the ratio of width of tubers to length 
in order to determine its value as an index to tuber form. Correlation 
studies for the ratio of width to length of tubers are presented in Tables 
IV to VII. The coefficients are here summarized. 
Table III .—Relation between width and length of tubers in the Early Ohio variety 
in igig 
Place grown. 
Coefficient of cor¬ 
relation. 
Number of 
individuals. 
Grand Rapids... 
0. 7i8±o. 014 
. 847± .025 
. 647 ± .032 
. 601 ± .034 
571 
316 
148 
161 
Duluth... 
University Farm, clay loam... 
University Farm, sandy loam. 
The data presented in these tables show that there is a decided corre¬ 
lation between the width and the length of tuber and that this correlation 
