SIZE OF POTATO SETS. 15 
In 1919 Selvig (So, p. 56-59) reported the results of size-of-seed 
studies made at the Minnesota Northwest Experiment Station, 
Crookstown, Minn., in which single-eye, 2-eye, and quarter- tuber 
sets each weighing an ounce, halved tubers 2 ounces in weight, and 
whole tubers weighing 2, 3, 4.1, and 9.5 ounces, respectively, were 
carefully studied with respect to the number of stems and tubers 
produced, The conclusions were that yields do not increase in direct 
proportion to the size of the set. Increased yields from the planting 
of halves and whole seed of the different sizes have not been sufficient 
to offset the increased cost arising from the extra seed needed and 
the necessity of hand planting. The number of stems increased 
slightly but gradually in sets of the same weight with increase in 
number of eyes on the set. 
In 1922 Rosa (78) reported the results of a four-year study of size 
of seed piece. The Early Ohio variety was used in this experiment, 
and four sizes of seed were studied in both the whole-tuber and cut- 
seed class. The tests were with seed pieces 20, 30, 40, and 50 grams 
in weight. Rosa's conclusions were that the most profitable size of 
seed piece to use in his section did not lie much beyond an ounce in 
weight, 
Salaman's conclusions (79) from a two-year study of the influence 
of size and character of seed on the yield of potatoes were as follows : 
The total yield varies directly with the weight of the tuber set. Small sets 
under 1 ounce in weight, although giving a greater return in proportion to their 
weight and a high proportion of heavy ware (large merchantable tubers) are 
uneconomical. Taking into consideration the total weight of seed used, the 
proportion of heavy ware produced, and the total yield, sets of 2 ounces in weight 
are the most remunerative. There is an inverse ratio between the size of the 
seed set and the percentage of heavy ware in the resultant crop. 
Studies of the results from whole small tubers and pieces of large 
tubers from the same plant by Stewart (92, p. 27) are summarized 
as follows : 
For seed purposes uncut tubers between 1 and 2 ounces in weight are at least 
as good as and probably a little better than pieces of equal weight cut from large 
tubers of the same plant . . . The yield of the potato plant increases as the 
number of stalks increases. 
The use of uncut tubers for seed does not always result in a large number of 
stalks or in a large proportion of small tubers in the crop. 
DEPARTMENTAL STUDIES OF THE SIZE OF THE SET. 
Experimental studies to determine the relative value of whole and 
cut seed and medium and small-sized tubers were undertaken by the 
Office of Horticultural and Pomological Investigations in 1913, but 
no carefully conducted tests were made until 1914. In that year 
studies were made at the Virginia Truck Experiment Station, Nor- 
folk, Va., and at Caribou, Me. In the following year the work in 
Maine was transferred to Aroostook Farm, Presque Isle, Me., and 
similar studies were undertaken at the potato experiment stations 
at Greeley, Colo., and Jerome, Idaho. The varieties experimented 
with at the four points mentioned consisted of the Irish Cobbler at 
Norfolk, Va., the Irish Cobbler and Green Mountain at Caribou 
and Presque Isle, Me., the Rural New Yorker No. 2 at Greeley, 
Colo., and the Charles Downing (Idaho Rural) and Russet Bur- 
bank (Netted Gem) at Jerome, Idaho. 
In the conduct of these studies considerable care was exercised 
to select tubers of uniform size, regardless of whether they were to 
