DEVELOPMENT OF TUBERS IN THE POTATO. 
21 
A study of Table VII shows that both the number and the weight 
of tubers per hill varied directly with the size of the seed piece 
planted. There was also a corresponding increase in the number of 
stems per hill. If, however, we consider the stem rather than the 
hill as the unit for comparison, we find that the number of tubers 
which set per stem was affected in only a very small degree by the 
size of the seed piece, though the general trend was toward a slight 
decrease in the number of tubers produced as the size of the seed 
increased. The effect oh the weight of the tubers per stem was 
much more pronounced, as there was a very consistent and well- 
defined decrease associated with increase in the size of the seed. 
It is apparent that the food supply was not sufficient to maintain 
the needs of the increased number of stems for maximum tuber 
production. 
Table VIII. — Relative influence of whole and cut seed on tuber production as shovm by 
the results ivith different kinds of seed where seed pieces of the same weight were used in 
tests of the Rural New Yorker variety of potatoes grown at Greeley, Colo., in 1916 and 
1917. 
Size and kind of seed. 
Weight 
of seed 
piece 
planted 
(ounces). 
Number 
of stems 
per hill. 
Number of tubers. 
Weight of tubers 
(grams). 
Per hill. 
Per stem. 
Per hill. 
Per stem. 
1 ounce, whole 
2 ounce, halved 
4 ounce, quartered 
2 ounce, whole 
4 ounce, halved ' 
1 f L76 
\ M L81 
j I 1.63 
\ „ 1/ 2.60 
/ l 1 2.50 
4.25 
3.99 
3.63 
6.31 
5.29 
7.18 
6.28 
4.79 
4.13 
2.41 
2.20 
2.23 
2.42 
2.12 
2.23 
2.06 
2.12 
2.11 
712. 72 
668. 61 
673. 25 
876. 55 
815. 51 
886. 23 
836.54 
764. 90 
737. 56 
404. 80 
368. 89 
413.20 
336. 74 
326. 65 
3 ounce, whole 
6 ounce, halved 
■ 
3 ounce, halved 
6 ounce, quartered 
} » 
/ 3. 22 
\ 3.05 
/ 2. 25 
\ 1.96 
274. 85 
274.14 
339. 32 
375. 95 
Comparisons of the results obtained from whole and cut seed 
recorded in Table VIII show a slightly larger number and, with one 
exception, a larger weight of tubers per stem where whole seed was 
used. 
RELATION OF VARIETY TO TUBER PRODUCTION. 
It is a matter of common observation that the tuber-producing 
ability of different varieties varies considerably with respect to the 
number and the weight of tubers per hill as well as other character- 
istics. This is brought out very clearly in the curves shown in 
figures 9 and 10. The data on which these are based were obtained 
in 1915 from 500 hills each of the Pearl and Rural New Yorker 
varieties taken from adjacent plats which were grown under condi- 
tions as nearly identical as possible. 
A study of the curves for the number of tubers (fig. 9) shows an 
entirely different distribution for the two varieties. In the Rural 
