Vermont Agricultural Experiment Station
Bulletin 177 (February, 1914)
LARGE SEED A FACTOR IN PLANT PRODUCTION
By M. B. CUMMINGS1

1It is a pleasure to acknowledge the writer's indebtedness to the station gardener, Mr. Stanley Hargreaves, who has given careful, continuous and intelligent service throughout the course of the trials and to whose conscientious devotion much of the success of the outcome is due.

Good seed, better seed, and the best seed are terms of prime importance in plant production. Good seed is that which is true to name, well preserved, viable and free of foreign seed and insect ravages. Better seed than that which is merely good may sometimes be procured by taking advantage of specially favorable seed years, and occasionally by the payment of higher prices for seed that is carefully screened. The best seed and the requirements therefor have not been fully determined. That parentage is one of the factors that makes for the best seed is a popular notion that is being supported by recent experimentation. Plant breeders are finding evidence of the value of parentage selection. It will soon be possible to formulate a law declaring that the best seed can be obtained by selecting it from plants which exhibit the greatest number of desirable characters, a point which will be discussed in part in this publication. The size of the seed is also a factor that determines its value; and the tests herein reviewed furnish evidence tending to show that, other things being equal, the largest seed are the best and the smallest seed the poorest.

The discussion of size of seed as a factor in plant production is not an entirely new idea. A review of literature shows that many isolated experiments have been conducted to determine the merits of large and small seed; but for the most part the work has been confined to the seed of farm rather than of horticultural crops. Good crop requirements, horticulturally speaking, are quite different from those called for in the culture of the cereals and grasses. Earliness, uniformity and high quality are imperative in horticulture, whereas with farm crops quantity seems to be paramount. Moreover, the intensiveness characteristic of horticultural operations emphasizes the significance of the use of the best seed.

The main items noted in this bulletin which relate to the importance of size of seed in plant production are: The influence of size on germination with special reference to rate, percentage, and uniformity; the effect of the size of seed on the size of the plant, involving a study of weight, height and general structure; the effect of size selection on flowering and fruiting habits, embracing observations on earliness and abundance of flowers and fruits, and on their quality; a study of the effect of continuous selection of large seed with special reference to general productiveness; investigations as to the cause for differences in size, and the place-formation of large and small seed; and, finally, a recognition of some economic considerations.

The source of seed used in these trials should be mentioned. Of necessity the initial plantings were made from sortings of commercial plantings; but thereafter, for the most part, the seed used were saved from our own cultures. In a number of cases the seed were obtained from individual plants which evidenced more or less striking peculiarities. This was especially observed in connection with beans, radishes, pumpkins, and squashes; the purpose being to compare size only, by eliminating the factor of mixed parentage, such as would inevitably enter in the case of seed purchased in commercial packages. The importance of this factor is discussed later (pages 100-101).

The seed were sorted and graded according to size, and, although the weights were recorded, the number of seed and performance of individuals were considered of special significance. The individual plant rather than the weight of the seed used, is the proper unit of comparison, although the results are sometimes better expressed in terms of percentages.

PREVIOUS WORK

1 U. S. Dept. Agr.. Year Book for 1890, p. 87.
2 Proc. Hort. & Vit. 30 (1899).
3 Univ. College Wales, Rpt, for 1899.
4 Amer. Breeders Asso. Rpts. 7, 8. p. 50 (1912).
5 Amer. Breeders Asso. Rpts. 7, 8. p. 61 (1912).
6 U. S. Dept. Agr., Bu. Plant Ind., Bul. 78 (1905).
7 Amer. Nat.. Vol. 44. No. 517, p. 48 (1910).
8 Amer. Breeders Asso. Rpts. 7, (1912).
9 Ont. Agr. Coll. Rpts. 28, 29, 32, 34, 35, ('02, '03, '06, '08, '09).

Experimental trials hitherto carried out as to the relative value of different sized seed have been of a somewhat fragmentary character, but the results attained have served to emphasize the importance of the subject from the standpoint both of theory and of practice. Considerable attention has been given to cereal seed. Hicks and Dabney1 state that large seed are heavier than small seed and germinate better. Degrully2 increased the yields of corn by one-third by the use of heavy seed, while oat yields were increased by 219 pounds grain and 303 pounds straw per acre. Middleton3 found with wheat and oats that large seed yielded nearly twice as much as did small seed. In studying correlation characters in oats, Leighty4 found that large kernels are as likely to occur in a head that produces relatively few as in one that produces many kernels. In observations made on fertility, variation and correlation of wheat, Myers5 found that the taller plants produce the heavier kernels and that an absolute correlation existed between the weight of grain and the weight of straw. Lyon6 has stated that both large and small grains in a lot of wheat must represent both large and small spikes, and that one is not necessarily selecting from the best plants if only the large grains are used. In correlation studies of heavy and light oat grains, Waldron7 reported that the heavy seed conies from the small rather than from the large plants. Love,8 however, takes exception to these statements and points out that if heavy or large wheat seed is used for planting it will come from the tallest and heaviest yielding plants. In a series of interesting experiments conducted by Zavitz9 on large plump and small plump seeds of barley, wheat, oats and field peas, it was found in every instance that the large seed produced the larger yield of grain.

10 Selsk, Khoz 1 Lyesow; 200 '01. Cited E. S. R 14. p. 432 (1902-3).
11 U. S. Dept. Agr.. Farm. Bul. 285 (1907).
12 U. S. Dept. Agr., Year Book for 1904. p. 435.
13 N. Y. (State) Sta.. Rpt. 8. p. 364 (1889).

A few other plants have also been studied in relation to large and small seed values. Experiments with sugar beets were made by Lubanski,10 who found both yield and quality increased by the use of large seed. Webber11 tested the merits of sorted heavy and unsorted light cottonseed and found substantial gains in favor of heavy seed, which was larger than the average unseparated seed. In trials with tobacco, Shamel12 separated seeds into three grades with respect to weight, and secured the best results with the heavier and the poorest outcome with the lighter seed. An interesting but incomplete study of beans by Beach13 indicates the superiority of large over small seed.

In view of the foregoing references which show interesting results among cereals and a few other plants, it seemed to be worth while to make tests with important horticultural crops; and to this end a series of experiments were conducted which are presented herewith.

SWEET PEAS

Since commercial sweet pea seed show considerable variation in size, the plant lends itself well to the purpose in hand; and it was decided to make productive capacity tests with large, medium and small seed of several varieties. The seed were sorted and graded by hand and their weights recorded. They were planted in parallel rows in good garden loam and, when several inches high, were thinned to equal distances so that all plants would be under comparable environment with respect to soil and air conditions.

As the plants advanced it was observed that those derived from the large seed were the most vigorous specimens, as shown by their greater height, more numerous branches, and more abundant foliage, a condition which did not obtain during the first 10 days of growth. However, the notetaking in this series of experiments was largely confined to blossom production.

Size
of
seed
Number
sown
Weight
of seed,
grams
Total
number
of plants
Total
number of
blossoms
Average number
of blossoms
per plant
Duke of Westminster
Large 50 4.98 45 11,821 263
Medium 50 3.70 40 10,069 252
Small 50 2.74 40 8,240 206
Earliest of All
Large 50 6.25 40 3,496 87
Medium 50 4.11 33 2,933 89
Small 50 2.51 16 567 35
Her Majesty
Large 50 4.50 34 12,244 360
Medium 50 3.19 35 10,585 321
Small 50 2.45 18 3,702 206
Coccinia
Large 50 5.13 40 16,266 407
Medium 50 3.07 40 13,578 339
Small 50 4.22 46 14,679 319
Agnes Eckford
Large 30 3.13 25 1,701 68
Medium 30 2.51 24 2,199 89
Small 30 1.87 25 2,145 82
Apple Blossom
Large 50 5.40 39 11,678 315
Medium 50 3.78 29 10,443 266
Small 50 2.65 38 9,592 273

The large seed presents several points of contrast with the small. It is heavier, germinates better, seems more likely to produce stronger plants, and, more important, yields a larger number of blossoms.

Earliness of blossoming is likewise to the credit of large seed. Records were made of the picking dates, thus enabling comparisons of relative earliness of flowering. A survey of the notes determined the decision to divide the sweet pea season into four periods, and, consequently, the results are grouped as noted below. The total number of flowers picked during each period, using 6 varieties, together with the percentage of the entire number for that period, is shown below for each size of seed.

EARLINESS OF BLOSSOMING, 1910

Variety and size
of seed
Number of
blossoms
Percent
of total
Number of
blossoms
Percent
of total
Number of
blossoms
Percent
of total
Number of
blossoms
Percent
of total
Totals
Agnes Eckford July 7 to 15 July 18 to 29 August 1 to 12 August 13 to 25  
Large 252 14.8 558 32.8 467 27.5 424 24.9 1,701*
Medium 136 6.2 693 31.5 322 14.6 1,048 47.7 2,199
Small 154 7.2 549 25.6 549 25.6 893 41.6 2,145
  6,045
Her Majesty July 9 to 23 July 26 to August 5 August 8 to 19 August 23 to September 3  
Large 1,019 8.3 2,909 23.8 3,717 30.3 4,599 37.6 12,244
Medium 669 6.3 2,181 20.6 2,660 25.1 5,075 48.0 10,585
Small 65 1.8 774 20.9 946 25.5 1,917 51.8 3,702
  26,531
Apple Blossom July 7 to 23 July 25 to August 5 August 8 to 19 August 23 to September 3  
Large 1,619 13.9 2,151 18.4 3,419 29.3 4,489 38.4 11,678
Medium 1,350 13.1 1,693 16.4 3,258 31.6 4,021 38.9 10,322
Small 961 10. 1,438 15. 2,862 29.8 4,331 45.2 9,592
  31,592
Coccinia July 11 to 23 July 25 to August 5 August 8 to 19 August 23 to September 3  
Large 2,183 13.4 3,196 19.7 4,982 30.6 5,905 37.1 16,266
Medium 1,857 13.7 2,415 17.8 4,265 31.4 5,867 36.3 13,578
Small 1,857 12.6 2,586 17.6 4,369 29.8 5,041 40. 14,679
  44,523
Earliest of All June 25 to July 7 July 9 to 29 August 1 to 12    
Large 484 13.8 2,468 70.6 544 15.6 ... ... 3,496
Medium 220 7.5 2,299 78.4 414 14.1 ... ... 2,933
Small 34 6. 267 47.1 266 46.9 ... ... 567
  6,996
Duke of Westminster July 7 to 20 July 25 to August 3 August 5 to 19 August 23 to September 3  
Large 1,010 8.5 2,417 20.4 3,825 32.4 4,569 38.7 11,821
Medium 602 6. 2,305 22.9 3,184 31.6 3,978 39.5 10,069
Large 805 9.8 1,790 21.7 2,313 28.1 3,332 40.4 8,240
  30,130
All Varieties  
Large 6,567 11.5 13,699 23.9 16,954 29.5 19,986 35.1 57,206
Medium 4,834 9.7 11,586 23.3 14,103 28.4 19,163 38.6 49,686
Small 3,876 9.5 7,404 19.2 11,305 29.1 16,340 42.2 38,925
  145,817
*Mildew.

A survey of this table indicates that in the season of 1910 the plants grown from large seed tended to bloom earlier than those grown from medium or small seed. This was in general true, whether measurements were made in terms of total numbers of early blossoms or in terms of percentage. With one exception, where mildew struck some of the large seed planted (Agnes Eckford), large seed plantings produced more total blossoms than did either medium or small plantings. Without exception they yielded more blossoms during the first picking period, usually 2 weeks long, and, save in the case of Agnes Eckford, the mildewed variety, more during the ensuing 10 days. The large seed plantings of four varieties furnished a higher percentage of their total blossoms in the first picking period than did the plantings from medium or small seed; and the proportions were as great in the other two varieties.

Stating the case another way, using the total figures at the foot of the table, there were 280 sweet pea seed of each size planted. Forty percent of the blossoms were derived from large seed plantings, 35 percent from medium sized seed plantings, and 25 percent from small seed plantings. The proportions of the blossoms from these three plantings picked early, in the midst of the season and late, were:

  Large seed Medium seed Small seed
Early 43.4 32. 24.6
Early midseason 41.9 35.5 22.6
Late midseason 40. 33.5 26.5
Late 36. 34. 30.

It would seem fair to assume that a sweet pea grower in a good season, such as was that of 1910, may reasonably expect more blossoms and more early blossoms if he grades his seed by size and chooses the larger seed than if he uses the average run of seed.

The experience of the summer of 1910 with sweet peas indicated the lower germinating capacity of small seed. Consequently it was planned to plant in 1911 a larger number of the small seed, so as to compensate for its poorer germination, and to secure after thinning a fairly uniform stand of plants. In any test of the productive capacities of plants it is necessary to afford comparable environmental conditions, such as distancing and the avoidance of shade, as well as care concerning fertilization, tillage, and training of the vines. Attention is drawn to these factors in order to explain the differences in the number of seed sown. However, as soon as assured stands, following germination, were in hand, a severe thinning process was carried out in order to insure as nearly as might be similar conditions for the plants grown from different sized seeds. Only from 2 to 6 percent of the original plants were left.

TOTAL AND AVERAGE BLOSSOM PRODUCTION FOR VARIETIES OF SWEET PEAS

Size of seed Number
sown
Number plants
after thinning
Total number
of blossoms
Average per plant
for the season
Earliest of All
Large 455 29 3,312 114
Medium 600 26 2,521 97
Small 1,240 24 2,331 97
Duke of Westminster
Large 432 25 490 20
Medium 600 22 319 15
Small 1,864 23 383 17
Her Majesty
Large 529 27 372 14
Medium 600 26 298 11
Small 319 25 373 15
Coccinia
Large 836 28 616 22
Medium 600 28 680 24
Small 656 21 442 21
Agnes Eckford
Large 450 29 406 14
Medium 600 28 265 9
Small 1,190 27 93 3
Apple Blossom
Large 304 18 284 16
Medium 610 18 183 10
Small 829 25 258 10
Gladys Unwin
Large 571 22 479 22
Medium 600 23 356 15
Small 1,330 22 392 18

The preceding table shows clearly the relative merits of large, medium and small seeds in blossom production. Importance is attached to the contrasts between the productive capabilities of plants from large and small seed. In every instance the plants grown from large seed were more productive than were those grown from small seed. The average gain in blossoms per plant of large as compared with small seed were: Earliest of All, 17; Duke of Westminster, 3; Coccinia, 1; Agnes Eckford, 11; Apple Blossom, 6; Gladys Unwin, 4.

The results attained with medium sized seed were variable, at times apparently less and at times more blossoms being secured than when small seed was planted.

Quality differences as between sweet pea blossoms are marked. First class blossoms have long and moderately stiff stems, with good sized blossoms, while second and third class blossoms can scarcely be used. Only in seasons of great scarcity would they be picked. All the 1911 blossoms were graded and each plant credited with its good and poor ones. At the close of the season it was found that the plants grown from large seed produced more good blossoms per plant than those grown from small seed, and, in every case but one, more per plant than those grown from the medium sized seed.

GOOD AND POOR BLOSSOMS FROM DIFFERENT SIZED SEED

Size of seed Number
plants in test
Total number
of blossoms
Number of
good blossoms
Number of
poor blossoms
Average number
of good blossoms
Average number
of poor blossoms
Earliest of All
Large 29 3,312 1,651 1,661 57 57
Medium 26 2,521 1,234 1,287 47 49
Small 24 2,331 1,248 1,083 52 45
Duke of Westminster
Large 25 490 360 130 14 5
Medium 22 319 261 58 12 3
Small 23 383 315 68 14 3
Her Majesty
Large 27 372 308 64 11 2
Medium 26 298 230 68 9 3
Small 25 373 271 102 11 4
Coccinia
Large 28 616 527 89 19 3
Medium 28 680 544 136 19 5
Small 21 442 331 111 16 5
Agnes Eckford
Large 29 406 329 77 11 3
Medium 28 265 221 44 8 2
Small 27 93 66 27 2 1
Apple Blossom
Large 18 284 185 99 10 6
Medium 18 183 132 51 7 3
Small 25 258 186 72 7 3
Gladys Unwin
Large 22 479 408 71 19 3
Medium 23 356 312 44 14 2
Small 22 392 336 56 15 3

The average gain in blossom production of good grade flowers from plants grown from large seed are: Earliest of All, 5; Duke of Westminster, 0; Her Majesty, 0; Coccinia, 3; Agnes Eckford, 9; Apple Blossom, 3; Gladys Unwin, 4. The gains with respect to blossom quality are less readily stated. It seems probable that a stricter grading system would have given more uniform results. Presumably other unconsidered factors may enter into this question, such as vigor of plants and parentage; yet on the basis of size the results are on the whole quite emphatic.

EARLINESS OF BLOSSOMING, 1911

Variety and
size of seed
Number of
blossoms
Percent
of total
Number of
blossoms
Percent
of total
Number of
blossoms
Percent
of total
Number of
blossoms
Percent of
total
Totals
Agnes Eckford July 24 to August 1 August 2 to 8 August 9 to 15 August 16 to 22  
Large 30 7.4 213 52.5 119 29.3 44 10.8 406
Medium 10 3.8 140 52.8 78 29.4 37 14. 265
Small 7 7.5 59 63.5 16 17.2 11 11.8 93
Coccinia July 18 to 30 July 31 to August 7 August 8 to 15 August 16 to 21  
Large 129 20.9 310 50.3 128 20.8 49 8. 616
Medium 85 12.5 392 57.7 124 18.2 79 11.6 680
Small 89 20.1 262 59.3 63 14.3 28 6.3 442
Earliest of All June 28 to July 13 July 14 to 25 July 26 to August 7 August 8 to 21  
Large 301 9.1 411 12.5 1,435 43.3 1,165 35.2 3,312
Medium 210 8.3 341 13.5 1,195 47.4 775 30.8 2,521
Small 129 5.5 311 13.4 1,082 46.4 809 34.7 2,331
Duke of Westminster July 15 to 26 July 26 to August 4 August 5 to 12 August 16 to 21  
Large 15 3.1 98 20. 226 46.1 151 30.8 490
Medium     43 13.5 184 57.7 92 28.8 319
Small 7 1.8 57 14.9 182 47.5 137 35.8 383
All Varieties  
Large 475 9.8 1,032 21.4 1,908 39.6 1,409 29.2 4,824
Medium 305 8.0 916 24.2 1,581 41.8 983 26. 3,785
Small 232 7.1 689 21.2 1,343 41.3 985 30.3 3,249
                  11,858
All Varieties (on basis of 100 plants each of large, medium, and small seed)
Large 428 9.8 930 21.4 1,719 39.6 1,269 29.2 4,346
Medium 293 8.1 879 24.2 1,520 42. 944 26. 3,636
Small 244 7.1 725 21.2 1,414 41.4 1,037 30.3 3,419
                  11,401

NOTE—The data for Her Majesty, Apple Blossom and Gladys Unwin are not presented owing to the bunching of the blossoming period within a comparatively short period. The meteorological condition in 1911 did not favor sweet pea growth.—

Meteorological conditions were perhaps not as favorable to sweet pea production in 1911 as in 1910; yet the general outcome as regards earliness of blossoming was identical with that observed in the previous season. The same results were attained as in the previous year; an earlier blooming tendency, more early blooms, larger proportion of early blooms, and more total blossoms were the results of using large seed. Thirty-eight percent of the blossoms were derived from the plants grown from large seed, 32 percent from the medium and 30 percent from the small seed; and the proportion of the blooms from the three plantings picked early, in midseason and late, ranged from 44 to 37 percent in the large seed plants, from 35 to 29 in the medium seed plants, and from 20 to 30 in the small seed plants. Clearly there is a distinct correlation between the earliness of bloom and the use of large sweet pea seed.

The earlier blossoming of plants grown from large seed may be in part explained by the greater strength and vigor of such plants when half grown. Comparative measurements were made in order to secure figures bearing on this point. One lot of plants was harvested and data collected relating to their weights and branching systems.

  Average weight
of plants, grams
Average number
of branches
Average length of
branches, inches
Large 11.6 4.7 11.5
Medium 7.3 3.4 11.
Small 6.2 3.7 11.5

The main stems and laterals of the plants grown from large seed were conspicuously larger and much more stocky, thus affording a condition or manner of growth of considerable importance. The slender spindling stems of plants grown from small seed are a limiting factor in flower production and undoubtedly represent relative lack of vigor or poor nutrition.

The practical aspects of these trials scarcely need review, but may be thus briefly summarized.

Early blossoms and first class flowers are more likely to be secured in greater number if the largest seed is used than if ungraded seed is sown. A relatively long sweet pea season, with a goodly number of early blossoms and a fair number during the remainder of the season, is more likely to follow the use of large than of ungraded seed.

A short blossoming season, with the bulk of the crop coming in midseason, is apt to be the result of the use of relatively small seed.

SWEET PUMPKINS

The considerable variation in the size of the seeds of sweet pumpkins occurring in individual fruits as well as in commercial packages suggested the advisability of making a comparative test of the merits of at least two sizes. An area of about one acre was devoted for 3 consecutive years to this one vegetable, with rows running alternately across the plantation, thus nullifying so far as possible the effect of local soil variations. The plants were injured during the first two seasons by long and severe droughts, yet results favored the use of large seed; hence a repetition of the trial was in order. Such ideal conditions prevailed during the third year, however, that the results were considered unquestionable.

Although the comparative studies were confined mostly to the fruits and fruiting season, it was observed that the small seeds germinated sooner, but even so, they did not produce as robust seedlings. Blossoming was practically simultaneous, the false or staminate blossoms coming on at the same time, whereas the true or pistillate blossoms were more abundant on the plants grown from large seed at the beginning of the blooming period.

The superiority of large seed is shown with respect to fruiting only in point of earliness. As the specimens began to ripen, counts were made of the number of well colored and ripened fruits. On September 3rd the large seed plants showed 8 fruits that were well yellowed, and 13 considerably pigmented, whereas those grown from small seed carried but two well colored fruits and only two others showing traces of yellow. On September 12th there were 50 fully ripe and 47 half ripe fruits on the large seed vines, and 21 fully ripe, and 45 half ripe fruits on the small seed vines. Specimens that were uniformly colored yellow were called ripe and ready for harvest, a judgment which was reinforced by a culinary test.

The differences in maturing are well shown by comparing figures at the time of harvest.

RELATIVE PRODUCTIVITY OF SWEET PUMPKINS WITH LARGE AND SMALL SEED

Size of seed Number
of vines
Number
of fruits
Total
weight, lbs.
Number of
ripe fruits
Weight
ripe, lbs.
Number of
fruits unripe
Weight
unripe, lbs.
Large 76 213 1,084 133 696 80 388
Small 76 256 1,247 68 386 188 861

The plants grown from large seed yielded 65 more ripe fruits, and 310 more pounds of edible product. The large seed plants averaged per vine 1 3/4 ripe fruit or 9 1/6 pounds of ripe fruit. The small seed plants averaged less than one ripe fruit per vine or a trifle over 5 pounds of ripe fruit. On the other hand the plants grown from small seed gave a greater total number of fruits, exceeding the large by 43 specimens, and a greater total weight, exceeding the large by 163 pounds, all of which were green at the normal time of harvest. Since the crop was not harvested until after the first hard frost, which occurred at about the usual time, and after nights had become so cool that the vines ceased to grow, it can scarcely be assumed that the plants from small seed showed greater productive capabilities. Under almost any cultural conditions, pumpkin vines will set more fruits than can be matured, and late settings invariably result in numerous small fruits. The superiority of the large seed is shown in its earliness, a matter of some theoretical interest as well as of much practical importance in northern districts or wherever early fruitage is desired.

Any practical test of the relative values of different sized pumpkin seed will be attended with difficulty if continued for more than a year, because of the fact that the flowers are unisexual. Foreign pollination by insects is likely to interfere with parental characters and values as well, so that the influence of seed selection might be entirely offset by the introduction of new factors.. The practice of hand pollination is also open to objection, since many failures result and delays in affecting the set are likely to vitiate results. However, the use of seed derived from a common parental fruit will eliminate much danger arising from irregularities of vigor.

HUBBARD SQUASH

Squash seed do not vary as widely in size as do those of many other plants, but if a large quantity is sorted sufficient differences in size and in number may be obtained to facilitate selection. The plantings were made under comparable conditions, the parallel hill-row' system being used. The results of the first year test are presented below:

Size of seed Number of
seeds planted
Weight of
seed, grams
Number
of vines1
Total number
of fruits
Total weight,
pounds
Number
salable
Weight Salable,
pounds
Large 150 48.9 75 111 1,330 110 1,324
Small 150 23.1 75 112 1,230 96 1,158
1 The plantings were thinned one-half after germination.

 It is interesting to note the points of contrast in large and small seed weights, and in the products of the different sized seeds. The large seed were more than twice as heavy as the small seed; yet they produced no more fruit in terms of number, although they yielded considerably more in terms of weight, furnishing an excess of 14 salable squash and 166 pounds more salable squash. There were 75 vines of each size of seed planted which would give a gain per plant of 2.2 pounds of edible product. The higher yielding capacity of large seed is not due to earlier germination, for the plants were nearly uniform in this respect. The seedlings derived from large seed were more stocky and hence it may be assumed that the production gain is attributable to the greater size of the embryos. A stronger growth force and greater vitality is therefore ascribed to plants which are grown from large seed.

To be of value, size selection of Hubbard squash seed should be done with care. Because of the unisexual nature of the blossoms crossings must occur, and obviously a mixed heritage introduces chaotic conditions. In order to insure proper seeds free from hybrid character, either hand pollination must be practiced or only seed derived from a common parental fruit used for sorting. When such precautions are taken, one may expect that large seed will do better than will smaller seed.

LETTUCE

Lettuce is a standard vegetable crop, largely grown in farm gardens and ranking high in commercial horticulture as a forcing crop; hence its inclusion in these seed sorting experiments. Although a seed-bed crop, it is often grown in places where space is expensive, where intensive culture prevails, and where crop uniformity and even maturity are of prime importance. On this account growers now carefully consider both seed and varietal choice. In some hothouse districts only certain varieties are deemed suitable for greenhouse culture; and a few progressive growers select only the locally grown seed now recognized as strains of commercial varieties. Although many methods of seed selection have been adopted, the influence of seed size has not hitherto been carefully studied.

Five varieties were used in these trials, namely, Hittinger's Belmont Forcing, Boston Forcing (hothouse varieties), Hanson, Deacon (garden kinds), and Iceberg (an excellent midsummer variety). This choice represents all seasons and methods of culture, and the likelihood of the general applicability of the outcome is thus enhanced.

Preliminary trials indicated that plants grown from large seed considerably outranked their competitors; hence an inquiry was made with a view of finding when and how these differences were manifested. Comparative observations were made of plants grown from large, medium and small seed in the seedling stage and at time of cutting. Five-week old plants furnished the following data:

MEASUREMENTS OF LETTUCE SEEDLINGS

Size of seed Height of plants
mm.
Length of leaves
mm.
Width of leaves
mm.
Boston Forcing
Large 25.9 16.8 13.4
Medium 24.1 15.6 13.2
Small 19.8 12.6 10.2
Hittinger's Belmont Forcing
Large 23. 13.1 9.7
Medium 21. 11.1 9.3
Small 18.2 9.9 8.9

The immature plants grown from the large seed have larger, longer and broader leaves. The difference was quite noticeable at germination and for nearly two weeks thereafter. In some cases, but not in all, this was true immediately after transplantation. In several instances conspicuous differences were less obvious than the actual measurements would indicate, a condition probably due to the abundance of leaves and to close formation. However, the differences reappear, as will be noted on the next page.

The relative gains for plants grown from large seed may be seen in the subjoined tables, the contrast being best stated in terms of averages.

Size of seed Number of
plants set
Number of
plants cut
Total weight,
grams
Average weight,
grams
Hanson
Large 33 33 12,960 392.7
Small 36 36 13,020 361.7
Deacon
Large 33 33 12,270 371.8
Small 8 8 2,580 322.5
Iceberg
Large 39 39 15,810 405.4
Small 32 32 12,690 396.6

The differences are small, but the uniform gain for large seed is interesting in showing constant and relative values. However, another basis of comparison must be used to get distinctive merits.

Since lettuce is seldom sold by weight, a comparison of the value of different sized seed on this basis is of little account. Uniformity of maturing and relative "heading-up" capabilities are of more importance. The formation of good, firm heads, making possible the growth of white, crisp, and highly edible center leaves, is characteristic of a good quality of lettuce. That good "heading up" characteristics are related to the size of the seed is indicated in the subjoined tabulation. The plants were cut when a majority of either group from sorted seed had formed mature heads, the sign of commercial and edible maturity.

SIZE OF SEED AND MATURITY OF HEADS

Size of seed Number
of plants
Number of
mature heads
Total weight,
  grams
Average weight,
grams
Percent
headed
Hittingers Belmont Forcing. (First trial)
Large 39 20 3,998 102.5 51.
Medium 40 18 3,997 99.9 45.
Small 38 11 3,087 81.2 29.
Hittinger's Belmont Forcing. (Second trial)
Large 30 20 3,495 116.5 67.
Medium 30 17 3,360 112.0 57.
Small 30 11 2,724 90.8 37.
Boston Forcing
Large 23 15 2,724 118.4 65.
Medium 24 12 2,623 109.3 50.
Small 24 9 2,169 90.4 38.

These figures accentuate the importance of large seed in lettuce production. The large seed plants formed from two-thirds to four-fifths as many more salable heads as (lid the small seed plants, and they were one-fourth to one-third larger. The results in each variety are concordant; the outcome is clean-cut. The intermediate sizes gave intermediate values.

The relative merits of different sized seed may be yet further noted by grading the plants that formed heads into two groups, those well headed and those poorly headed; and by grading those that failed to head at the normal time into good and poor specimens. The last subsection includes the plants that are small, irregular in shape, and loose in leaf formation. The tabulation below shows the distribution into the groups, and emphasizes the superiority of large seed.

Size of seed Number
of plants
Number headed Number not headed Weight,
grams
Average
weight,
grams
Well Poorly Good
specimens
Poor
specimens
Hittinger's Belmont
Large 159 68 28 44 19 26,876 170
Small 153 50 39 34 30 24,547 160
Difference in favor of large   18 11 10 11 2,329 10

The plants grown from large seed headed up better and were heavier than those grown from small seed. Sixty percent of the former headed up, 56 percent of the latter; 43 percent of the former gave good heads, 33 percent of the latter.

Another trial afforded similar but less emphatic results as shown on the next page.

Size of seed Number
of plants
Number headed Number
not headed
Total weight,
grams
Average weight,
grams
Well Poorly
Hittinger's Belmont
Large 90 54 9 27 11,906 132
Small 90 48 14 28 10,992 122
Difference   6 5 1 914 10

In this instance the large seed plants gave better results by exhibiting, on the whole, better formed heads—a result not readily made manifest in tabular form—and by producing heavier plants.

Several points are worthy of note in summarizing. Marked differences in favor of large seed appear in the seedling stage, a point of little value in itself were it not for the fact that an early advantage influences later growth. Large seeds start the plants off better; and greater leaf surface area in early life is of permanent benefit. Moreover heavier plants, better heads, and greater uniformity at edible maturity are usually secured. In every instance and at almost every stage of growth it could be seen that the plants grown from large seed were much more uniform in stature and in time and manner of heading. Plants grown from small seed were very variable in size and quality-some very good, a few mediocre, and many very poor. Some headed early, but most of them were tardy in forming the heart and in firmly filling the head.

It seems reasonable to conclude that a large sized seed is a factor in producing head lettuce of good quality and earliness of maturity. In the writer's judgment the lack of plant uniformity commonly observed in commercial lettuce culture is quite apt to be due to the use of seed which is variable in size and, consequently, variable in value.

BEANS

Many trials extending over three seasons were made with beans. The plants were grown in the greenhouse and in the garden. All varieties of beans exhibit variation in size of seed, but some kinds afford greater contrasts than do others. Naturally choice was made of such desirable commercial sorts as were relatively variable in this particular. Preliminary tests were made with Golden Wax, a variety which shows conspicuous differences in size among unsorted beans. Numbers and weights were taken into account, but especial attention was paid to numbers, so that the productive capacities of individuals could be compared. Plantings of the sorted seeds were made in parallel rows so that comparable cultural conditions would prevail. A serious midsummer drought accounts for the very low yield and the large proportion of empty pods.

GOLDEN WAX—SEASON OF 1910
GREEN WEIGHTS

Size of seed Number used Weight of seed,
grams
Number of
plants
Weight of
green pods,
grams
Average weight
of pods per plant,
grams
Large 500 312 426 5,894 13.8
Medium 500 252 429 4,762 11.1
Small 500 169 370 3,629 9.8

 AIR DRY WEIGHTS

Size of seed Dry weight of
pods, grams
Total number
of pods
Number of
empty pods
Number of pods
containing one
or more beans
Number
of beans
Weight of
beans, grams
Large 1,899 2,545 1,580 965 2,503 964
Medium 1,389 2,032 1,390 642 1,578 624
Small 709 1,211 694 517 1,445 567

The large seed produced a larger number of more thrifty plants, more green and more dry pods, and more beans both in terms of number and of weight. The plants grown from small seed set fewer pods by 52 percent. No material differences occur in the number of beans to the pod or in the weight of the beans grown from the different sized seed.

Every bean grower knows that individuals vary greatly in productivity, for reasons either unknown or not well understood. Size of seed undoubtedly influences yields and past productivity may also be a factor. A comparative test of these two factors was made possible by keeping the yields of different plants separate at harvest, and then sorting the seeds with reference to past productivity and present size.

SIZE OF SEED, 1911

Size of seed Number of plants Number
of pods
Average number of pods per plant Average weight of pods per plant, grams
Large 470 3,405 7.24 28.3
Medium 431 2,535 5.87 26.0
Small 574 3,535 6.16 25.8

A gain of 1.08 pods per plant is worth while if multiplied by the number of plants per acre. It is a sixth increase. And yet more is a gain of 1 3/8 pods, or nearly one-fourth increase.

In the tables on the last page, note is taken of size only; but this test was conducted from two points of view, namely, size and parentage. The term "parentage" as thus used refers merely to the productive capabilities of this lot of plants in the previous year. At several harvest times the product of each plant was kept separate, so that eventually two groups were formed, composed of high and low yielders. Plants that bore over 30 seeds were classed as high yielders, and those bearing less than 10 seeds were classed as low in yield. The object in view was to endeavor to determine if selection according to past productivity was of more value than was selection with reference to the size of the seed.

Size and former size. The former size as well as the present size of seeds seems to be a factor in production. The figures given below epitomize the results attained from extensive plantings in 1912, and show the relative importance of giving attention to the size-parentage of seeds.

  Number
of plants
Number
of pods
Total weight
of pods, grams
Total weight
of shelled
beans, grams
Weight of
empty pods,
grams
Production per
vine, pods and
beans, grams
Production per
vine, shelled
beans, grams
Average number
of pods per vine
Large from large 1,576 11,503 31,015 22,424 8,591 19.7 14.2 7.3
Large from small 1,650 10,674 28,993 21,601 7,392 17.6 13.1 6.5
Small from small 2,587 11,270 29,930 21,770 8,160 11.6 8.4 4.4
Small from large 1,899 11,358 30,906 22,452 8,454 16.3 11.8 6.0

The number of plants grown from the various groups differs on account of difficulties encountered in thinning and spacing. The averages show the largest yields when large seed derived from plants grown from large seed were planted, and the smallest yields when small seed derived from plants grown from small seed was used. It is interesting to note that the yields derived from small seed produced from plants grown from large seed were better than when the small seed came of small seed ancestry, thus indicating the likelihood of lessening yields from the successive planting of small seed. Furthermore, large seed derived from plants grown from small seed outclassed small seed which was the offspring of plants grown from large seed, a triumph of individuality over heredity. In view of large number of plants in the test, the great care used in sorting the seed, and the uniform cultural conditions which obtained, the writer deems the outcome to have been a representative one.


PLATE I—Sweet peas grown from large seed (left), small seed (middle), and medium sized seed (right).


PLATE II—(1) Spinach crop from large (left) and small (right) seed. (2 and 3) Alternate rows
from large and small seed. (4) One hundred beans, before and after grading as to size.


PLATE III—(1) Radishes from medium, small and large seed (100 seed of each). (2) Lettuce;
alternate rows from large and small seed. (3) Seedling Hubbard squash from large, small and medium seed.


PLATE IV—Golden wax beans, September 3rd, grown from large seed (left), small seed (middle), and medium sized seed (right)

The results attained with other varieties in 1912 are tabulated below. The outcome was identical with that secured with Golden Wax (page 105) emphasizing the superiority of large seed.

Size of seed

Number
of plants

Number
of pods

Weight of
pods, grams

Weight shelled
beans, grams

Weight of empty
pods, grams

Pod production
per vine, grams

Shelled beans
production
per vine, grams

Average number
pods per vine

Horticultural Dwarf
Large 414 3,536 10,995 7,655 3,340 26.6 18.5 8.5
Small 482 3,528 9,875 6,577 3,298 20.5 13.7 7.3
Brittle Wax
Large 541 11,557 29.092 20,610 8,482 53.8 38.1 21.4
Small 451 8,594 18.955 14,742 4,213 42. 32.7 19.1
Stringless Green Pod
Large 3,226 22,177 60,011 44,027 15,984 18.6 13.7 6.9
Small 4,486 22,268 60,842 44,226 16,616 13.6 9.9 4.9

So far as beans were concerned attention was directed chiefly to three points; earliness of production, amount of production, and the relative importance of size of seed and of past productivity. With respect to earliness of production, it was shown with Golden Wax that in comparison with the medium seed, large seed is early and small seed late. Many plants grown from small seed were so late in maturing that they were injured by frost. In comparing past productivity and present size, it appeared that the basis of selection influences the yield, that heredity is probably of more importance than is size, and that neither heredity nor size can be reckoned with alone without recognizing the influence of the other factor. The amount of production either of full or empty pods shows that there is much advantage in the use of large seed; and this statement holds whether one considers string beans or seed beans.

THE EFFECT OF PAST PRODUCTIVITY

Previous
production
Number
of plants
Number
of pods
Weight of
pods, grams
Average number
of pods per plant
Average weight of pods
per plant, grams
High yielders 617 4,260 20,140 6.9 32.7
Low yielders 854 5,212 19,730 6.1 23.

The averages tell the story. The high yielders gave an average of 0.8 of a pod more per plant than did the low yielders; in terms of weight the average pod produced by high yielders exceed the low in weight by nearly 10 grams (1/3 ounce), which is a pronounced gain.

This initial trial would seem to indicate that with Golden Wax, size of seed and parentage of seed possess nearly equal values in determining productivity, and that selection on either basis will be likely to produce a marked gain, which, in terms of acre yields, means a largely increased income.

PARSLEY

This biennial herb represents a different group of plants than the others used in these trials. In view of the rapid renewal of its top which permits frequent cuttings, it seemed especially adapted to this study. The moss curl variety most commonly grown in Vermont was used. The test was conducted in the greenhouse under conditions of perfect control. The plants from large, medium and small seed were planted in alternating rows in good soil on a greenhouse bench on December 1, and cuttings were made on February 28, March 22, and May 6.

FIRST CUTTING, FEBRUARY 28

  Number
of plants
Total
number
of leaves
Total
weight,
grams
Average
number
of leaves
Average
weight per
plant, grams
Average
weight per
leaf, grams
Large 66 592 1,270 9.0 19.2 2.1
Medium 67 540 925 8.1 13.8 1.7
Small 64 404 594 6.3 9.3 1.5

 The first cutting was delayed until the plants grown from the small seed were sufficiently large to warrant a harvest. The plants grown from large seed were clearly ready for a cutting a week earlier than those grown from the small seed. The superiority of large over small seed is shown in that 2.7 more leaves per plant were grown, an increase of 43 percent, as well as larger and heavier leaves, weighing more than twice as much. And the same tale is told in the second and third cuttings. The smaller number of plants grown from small seed in the first and second cuttings was due to the failure of a few plants to produce enough for cutting. There were also two plants grown from medium seed that failed to produce leaves for the first two cuttings.

SECOND CUTTING, MARCH 22

  Number
of plants
Total
number
of leaves
Total
weight,
grams
Average
number
of leaves
Average
weight per
plant, grams
Average
weight per
leaf, grams
Large 66 317 599 4.8 9.1 1.9
Medium 67 333 483 5.0 7.2 1.5
Small 59 230 299 3.9 5.1 1.3

THIRD CUTTING, MAY 6

  Number
of plants
Total
number
of leaves
Total
weight,
grams
Average
number
of leaves
Average
weight per
plant, grams
Average
weight per
leaf, grams
Large 66 497 709 7.5 10.7 1.4
Medium 69 497 643 7.2 9.3 1.3
Small 65 407 447 6.3 6.9 1.1

 Large parsley seed is superior in that its plants produce heavier and earlier foliage crops, more leaves per plant, and continue to be more vigorous even after two cuttings have been made. This continued superiority suggests a better root system, which one might expect to be associated with a larger shoot system. The greater recuperative capacity of plants grown from relatively large seed is a factor of considerable importance in parsley culture.

RADISH

The radish was selected partly because its seed exhibit much variation in size and weight, but chiefly because it it one of the shortest of the short-term crops. Sixteen different trials were made during a period of 3 years. Only two varieties were used, but the many trials made have afforded fairly uniform and consistent results. Observations made on seeds of other varieties show the same divergencies in size of crop and there seems no reason to suppose that the varieties chosen were in any way abnormal.

Attention was first directed to the general results of seed selection with reference to size. The relative values of large and small seed is shown in the tables presented below:

Size of seed Number
of plants
Weight
of roots,
grams
Average
weights,
grams
Number ones Number twos
Number Weight,
grams
Number Weight,
grams
Early French Breakfast
Large 78 890 11.4 32 510 46 380
Small 55 425 7.7 4 50 51 375
Early Scarlet Globe
Large 53 765 14.4 23 440 30 325
Small 56 573 10.2 16 220 40 353

The superiority of large seed is shown in the greater weight of the edible portion of plants, approximately a 50 percent increase; and in the greater proportion of number one plants, both in terms of numbers and weight. Forty percent of the crop grown from large seed was classified as number one, while only from 7 to 28 percent of the plants grown from small seed were thus classified.

After noting the relative productive capabilities of large and small seed, it then seemed necessary to observe other factors in comparing the different sized seeds. The following table displays data as to variations in weight of seeds, number of plants produced, together with the quality and weight of plants at the time of harvest.

Size of seed Number
sown
Weight
of seed, grams
Number
of plants produced
Germination
percent
Total weight
of roots at
harvest, grams
Average
weight at
harvest, grams
Number
edible
Percent
edible
Weight
edible,
grams
Number
not edible
Weight
of tops,
grams
Large 4,700 54.4 3,466 74 46,560 13.4 2,688 77.5 21,120 778 23,040
Medium 4,700 43.5 3,270 70 36,960 11.3 2,077 63.5 14,880 1,193 14,400
Small 4,700 24.7 2,316 49 21,960 9.5 1,325 57.2 10,560 991 9,600

The large seed outweighed the small by 120 percent; its viability was better by 50 percent; the crop from the same number of seed grown was 117 percent larger in terms of weights, 104 percent larger in terms of numbers, 100 percent greater in terms of weight of edible roots, and the tops were 135 percent heavier. The total edible weight of the crop secured where small seed was used was just 50 percent of that harvested from the sowing of large seed.

The studies noted above were followed by others at germination time. Weights were made of the embryos during the first 5 days after applying moisture to the seeds. The differences in weight are shown in the table below:

WEIGHT OF EMBRYOS
RELATIVE WEIGHTS OF 15 SEEDS DURING GERMINATION

Size of seed Weight of seed, grams First day, grams Second day, grams Third day, grams Fourth day, grams Fifth day, grams
Large .477 .654 .885 1.109 1.127 1.307
Medium .432 .484 .560 .... .849 1.042
Small .316 .411 .434 .462 .581 .703

The greater weight of the embryos of the large seed should be associated with the greater weight of mature plants, as noted in a preceding table. It seemed also desirable to correlate the matter of greater weight with that of the relative sizes of embryos; consequently there is added another table showing their measurements.

SIZE OF EMBRYOS

Days after
germination
Number of
measurements
Large Medium Small
Width of
leaves, mm.
Length of
embryo, mm
Width of
leaves, mm.
Length of
embryo, mm
Width of
leaves, mm.
Length of
embryo, mm
1 15 7.3 25 6.9 24. 5.6 21.
2 16 7.4 33 6.6 27. 5.5 21.
3 15 7.3 41 6.3 32. 5.3 21.
4 30 7.4 43 6.8 36. 5.5 27.
5 30 7.7 47 7.2 39. 5.9 32.

Beginning with the first day after germination and continuing for five days it was found that the embryos of large seeds have wider and longer leaves than those of small seeds, while the embryos of medium sized seeds are dimensionally intermediate. If plants are larger from their start and even before germination, it is to be expected that they will continue to prove superior in size as they develop.

Observation has shown that radishes grown from large seed attain the edible stage sooner than those grown from small seed. It was thought that on this account they might contain more water; hence a number of plants were trimmed as for eating and their dry matter contents determined:

Size of seed Weight before
drying, grams
Weight after
drying, grams
Water
percent
Dry matter
percent
Large 43.8 1.84 95.79 4.21
Medium 29.2 1.55 94.70 5.30
Small 31.3 1.59 94.94 5.06

This table is representative of several similar and concordant determinations made but not published. It shows the lower dry matter content of plants grown from large seed. The plants were of the same age from seed, but were not of the same degree of maturity. Plants grown from small seed were about a week later in coming to edible maturity than those grown from large seed. It is doubtful if an exact comparison could be made by allowing the small seed crop to grow another week, for by so doing changed soil and air conditions might intervene. However, the water content of a radish is not a matter to cavil about. It is always extremely high. Radishes are not eaten for the food they contain, but as a relish. Good quality radishes must be rapidly grown, crisp and succulent, no matter how high their water content. Aside from size as related to quality, the plants grown from small seed were as good as those grown from the large seed.

The source of seed, whether from one parent or several, seems to make little if any difference in the value of the product. The following tables show the results of seed sorting according to size when the seeds in a series are derived from the same plants, the purpose in this instance being to eliminate the influence of individuality in different plants, and to insure similarity in such parental factors as relate to vigor and general growth force.

RADISH SEED OF DIFFERENT SIZES DERIVED FROM THE SAME PARENT

Size of seed Number
of plants
Total weight
of plants,
grams
Average
weight
of plants,
grams
Weight of
tops, grams
Average
weight
of tops,
grams
Weight of
roots,
grams
Average
weight of
roots,
grams
Actual gain of
large over small
seed in weight
of roots, grams
Production gain of
large over small
seed in weight
of roots
Planted December 30, harvested March 1
Large 39 737 18.9 335 8.6 402 10.3 1.7 20
Medium 40 678 17.0 307 7.7 371 9.3 .. ..
Small 36 549 15.3 241 6.7 308 8.6 .. ..
Planted January 6, harvested March 1
Large 39 475 12.2 280 7.2 195 5. 1.3 35
Medium 44 502 11.4 292 6.6 210 4.8 .. ..
Small 38 372 9.8 233 6.1 139 3.7 .. ..
Planted December 30, harvested March 1
Large 100 1,658 16.6 777 7.8 881 8.8 2.3 35
Medium .107 1,503 14.1 713 6.7 790 7.4 .. ..
Small 103 1,248 12.1 577 5.5 671 6.5 .. ..

Fully as emphatic results follow the selection of seed all of which is from the same plant, as were secured by selection on the size basis irrespective of parentage; so that the use of commercial seeds of large size has practical bearing irrespective of considerations of parental vigor.

Large seed regardless of its source, of mixed or unmixed heritage, is superior to small seed of the same source, because it gives larger plants, greater uniformity in stand at edible maturity, and a maturation gain of from 7 to 10 days. In greenhouse culture and elsewhere when space is valuable or earliness imperative, an economic gain may be expected from the use of large seed. In actual practice this advantage may be secured by sifting out and discarding the small seed.

SPINACH

Commercial spinach seed show marked variations in size and weight. Repeated trials with seeds sorted according to size showed that large seed weigh from three to four times as much as do the small seed. Therefore this plant afforded an exceptional opportunity to test out the productive capacities of large and small seed. Seven tests were made and, with one exception, the plants from large seed were superior in every respect.

Two varieties known in the trade as Victoria and Long Standing, seemed best adapted to the purpose. The results of these trials appear herewith:

Size of seed Number
planted
Weight of seed,
grams
Number
germinated
Germination
percent
Total weight of
plants grown, grams
Average weight
per plant, grams
Average number
of leaves per plant
Average vertical
diameter of leaves,
millimeters
Average transverse
diameter of leaves,
millimeters
Long Standing
Large 200 2.73 121 61 472 3.9 13 43.4 35.7
Medium 200 0.81 116 58 293 2.5 11 37.8 32.0
Small 200 0.70 53 27 110 2.1 9 33.6 25.9

The medium sized seed intermediates with reference to weight of seed, germination, weight of plants, number of leaves, and average dimensions of leaves, and obviously the large outclasses the small.

Another trial with the same variety gave similar results:

Size of seed Number
sown
Weight of
seed, grams
Number
germinated
Germination
percent
Total weight
of plants, grams
Average weight
per plant, grams
Average number
of leaves per plant
Long Standing
Large 300 3.83 151 50 709 4.7 11
Medium 300 1.71 136 45 571 4.2 10
Small 300 1.02 95 32 405 4.3 9

Progressive harvesting raises another point as to relative seed values, viz.: the proportion of the crop cut at a given time. Long Standing was cut at several different dates, as soon as the plants began to form flower stalks and all leaves an inch or more in length were counted, with results as follows:

Large seed Small seed
Date of
cutting
Number
of plants
Weight of
plants, grams
Number
of leaves
Plants cut,
percent
Number
of plants
Weight of
plants, grams
Number
of leaves
Plants cut,
percent
Long Standing
May 3 25 734 386 12 20 321 253 10
May 7 64 1,928 1,093 31 40 1,039 653 19
May 10 29 1,150 559 14 20 484 193 10
May 17 78 2,451 1,362 38 103 3,099 1,753 50
May 20 9 337 175 4 22 864 387 11
Totals 205 6,600 3,575 100 205 5,807 3,239 100

The plants grown from the large seed outstripped those grown from the small seed in every particular, though the differences are not extensive. They are advantaged in that they are earlier, heavier and more leafy plants. With respect to earliness they matured 58 percent of the crop by the third cutting, at a time when the small seed plants had matured only 39 percent. With respect to total yield they outclassed their competitors by 14 percent. The plants grown from large seed carried 10 percent more leaves.

In another trial with Long Standing spinach, note was made as to the relative heights of the plants which were grown in a cold frame during late spring and early summer, and allowed to blossom and set seed but not fully to mature.

Size of seed Number
sown
Number of
plants
germinated
Germination
percent
Average
weight
plants, grams
Average
number
of leaves
Average height
of main flower and
stalk, millimeters
Large 200 190 95 84.7 31.5 148
Small 200 168 84 60.8 25.2 119
In favor of large   22   23.9 6.3 29

A further trial of Victoria was made, the crop being cut 65 days after planting.

Size of seed Number
sown
Weight of
seed, grams
Number of
plants produced
Germination
percent
Total weight,
grams
Average weight
of plants, grams
Large 288 4.2 208 72 757 3.6
Medium 288 3.1 156 54 547 3.5
Small 288 1.6 112 39 252 2.3

These trials were conducted in various places, in the garden, cold-frame and greenhouse, and doubtless represent such results as might reasonably be expected under suitable commercial conditions.

GARDEN PEAS

Garden peas were tested with reference to earliness of pod production, quantity of pods produced, and the amount of edible peas. The Bliss Everbearing variety was used, and equal numbers of seed of the several sizes were secured by sorting a commercial purchase into groups. The plants were grown in alternating rows with a view of securing fairly comparable conditions of growth and development. Periodical observations were made with respect to growth. But few differences occurred in the germination of the several lots, although it was clear that the large seeds produced somewhat stronger plants with a greater tendency to form branches. The plants seemed to blossom with about the same profusion after the first week of flower production, but with the first count for blossoms the large seed showed a marked gain for earliness, the figures being, large, 988; medium, 755; small, 483; a result in complete harmony with that obtained with sweet peas (page 92).

The pods were picked as fast as they reached marketable size with results as follows:

Size of seed Number of vines Weight of peas and pods, grams Number of pods Weight of peas, grams Weight of pods, grams Average number of pods per vine Average weight of pods per vine Weight of peas per vine, grams Average weight of empty pods per vine, grams
Bliss Everbearing
Large 1,758 83,490 16,224 25,449 8,041 9.2 47.5 14.5 33.
Medium 1,954 80,628 17,760 25,249 55,379 9.1 41.3 12.9 28.3
Small 1,823 78,064 19,201 26,787 51,277 10.5 42.8 14.7 28.1

These figures are interesting. They are not in accord with those secured using other crops. The plants grown from small seed yielded more but lighter pods than those grown from large seed, yet they contain a slightly greater weight of green peas. Plants grown from the large seed lead those grown from the small seed only in two particulars; they produce more early blossoms and heavier pods, but they do not yield as many edible peas. However, peas are usually marketed in the pod and sold by volume rather than by weight; hence the size of the pods has a commercial bearing. Assuming the correctness of this outcome and its general applicability, the producer advantages slightly by the use of large seed; but the buyer does not, since the extra volume and weight secured by the use of the large seed represent pods and not peas. The differences in production are better shown when the plants were allowed to mature and the seeds to ripen.

COMPARISONS WITH RIPE SEEDS

Size of seed Number
of vines
Number
of pods
Total weight,
grams
Weight of peas,
grams
Weight of empty
pods, grams
Average number of
pods per plant
Production per
vine, grams
Production per vine
shelled peas, grams
Large 829 7,870 10,546 8,529 2,017 9.5 12.7 10.3
Medium 993 8,439 10,822 8,799 2,023 8.5 10.9 8.9
Small 830 9,122 10,138 8,231 1,907 11.0 12.2 9.9

The plants from small seed produce more pods by count, but they are lighter by weight both with and without seeds, thus showing a slight advantage in the use of the large seed. The medium size is utterly outclassed. The greater productivity of plants where large seed was used, a 4 percent gain of dry seed, may be due to their greater vigor and larger size. The assembled figures bearing on this point are presented herewith:

AVERAGE OF PLANT MEASUREMENTS

Size of seed Weight of
plant, grams
Number
of leaves
Number
of branches
Length of stem
and branches, mm.
Large 11.6 37. 4.7 114.7
Medium 7.3 41.2 3.2 115.0
Small 6.2 44.5 3.7 115.2

Small seed produced much lighter plants with fewer, and necessarily more slender, branches. The large seed plants carried a less number of leaves than did the small seed plants, but these leaves were heavier, darker green in color, thicker, and presumedly as a consequence had better assimilative organs. In short, plants grown from large seed are the more robust.

If one may judge from these single trials the garden pea seems an anomaly in that, unlike the other plants under trial here and elsewhere, the size factor does not appear to be important in so far as the early increase of its edible crop is concerned. The number of trials, however, was not large enough to warrant one in dogmatizing.

PLACE ORIGIN OF SMALL SEED

In connection with the study of the value of different sized seed, the question was raised as to the place origin of small seed. The collection of data bearing on this point was confined to a few plants, and is introduced here merely as a matter of interest.

Observations on beans show considerable regularity with respect to the placement of small seed in the pods. An examination of 3,000 pods showed that in 18 percent of the cases the small seeds occur in the middle of the pod, in 33 percent in the distal end, in 49 percent in the basal end. The ends of the pods contribute most abundantly to the total quantity of small seed. When only a single seed occurs in a pod—as is rarely the case—it is neither necessarily small nor located in any particular position within the pod.

The placement of small pea seeds follows a definite plan. Almost without exception the small seed occurs at the end of the pod; and observations on hundreds of pods shows that the distal end is as prolific of small seed as is the basal end. This is exactly what one would expect from a theoretical point of view, for the nature of the ovary is such that the distal ovules would be the last to be reached by pollen tubes and the basal ones would be the first to be fertilized, with the central ones intermediate. A detailed study of this point would lead us far from the general theme of these trials. However, priority of fertilization may not be the determinative factor. All the seeds, small as well as large, which develop to any extent have been fertilized, and apparently some selective process is at work which determines what seeds are to be the best developed. All the ovules may be fertilized, yet this does not guarantee development, or the equal nutrition of all the seeds in a pod. It is a fact of common observation that unfertilized blossoms are first to fall, but familiar examples are evident on every hand of wild and cultivated plants which shed many fertile blossoms. Unfavorable conditions levy a heavy tax, and the first evidences of their effect are known to the laymen as "blossom blight" and "June drop." If the first shedding is insufficient, another may occur; and, finally, close to the maturing stage of the plant and its seed, some seeds, though ripe and matured, are merely undersized. This explanation seems the more reasonable when it is realized that the size of a seed is as much a matter of the size of its embryo as of that of its food supply.

COMPARISON OF PLANTS AT DIFFERENT AGES

A point of interest in explaining the variations in plant production is that of the relative sizes and weights of plants grown from small and from large seed at different stages of growth. Conspicuous divergencies begin to appear in the size of their leaves at periods of 10 and 15 days after germination. It was to be expected that the first assimilation leaves which develop from the plumule would be larger in plants grown from large seed than in those grown from small seed; and such is the fact; but other and later leaves still more strongly testified in size and weight to the superiority of large seed. Visible and apparent differences were reinforced by comparative measurements. The data showing weights and surface areas of the foliage portions of beans are presented herewith:

COMPARISON AFTER FORTY DAYS' GROWTH

Size of seed Large Medium Small Percent gain
of large
over small
Total weight four plants 56.9 grams 39.1 grams 35.7 grams 59
Average weight per plant 14.2 grams 9.0 grams 8.9 grams ..
Total weight eight leaves 30.0 grams 23.2 grams 21.0 grams 46
Average weight per leaf 3.8 grams 2.9 grams 2.6 grams ..
Total surface area eight leaves 197.2 sq. in. 139.3 sq. in. 129.9 sq. in. ..
Average surface area per leaf 24.5 sq.in. 17.4 sq.in. 16.2 sq. in. ..

These figures show clearly the relative sizes of plants produced from the different sized seed. The superiority of large over small seed is shown in the greater weight of the whole plant, in the greater weight of leaves and the leaf surface area. The medium sized seeds gave intermediate results.

Comparisons were made at flowering time, which was 60 days after planting, with the contrast in terms of average weight and number of leaves as presented below:

COMPARISON AFTER SIXTY DAYS' GROWTH

Size of seed Number
of plants
Weight of
plants, grams
Average
weight, grams
Total number
of leaves
Average numbe
r of leaves
Large 31 2,340 75.5 981 31.6
Small 32 1,786 55.8 767 24.0

The large seed plants outweighed the small seed plants by 31 percent, and their leaf production exceeded that of their competitors by 32 percent. The increased weight may be an expression of greater vigor, and more leaves doubtless means more assimilation capacity, factors which should have bearing on pod and seed production.

The next step was to compare the plants at the time when the pods were well developed, but before the seeds were large enough to form conspicuous knobs. The plants were then 90 days old. No pods less than 2 inches long were counted.

COMPARISON AFTER NINETY DAYS' GROWTH

Size of seed Number of vines Weight of vines, grams Number of pods, grams Weight of pods, grams Average weight of vine, grams Average number of pods per vine Average weight of pods per vine, grams
Large 30 3,340 175 1,243 111.3 5.8 41.4
Medium 28 3,067 174 1,139 109.5 6.2 40.7
Small 31 2,529 120 639 81.6 3.9 20.6

The relative merits of the two sizes of seed are well exhibited in this table. The plants grown from large seed were three-eighths heavier, and produced more pods which weighed twice as much per vine. The medium seed yielded much better plants than the small and almost as good plants as did the large seed.

Plants grown from large seed being larger, heavier, leafier. possess distinct advantage in all stages of plant activity over their less well endowed competitors.

STRUCTURAL DIFFERENCES IN EMBRYOS

The relative sizes of embryos of large and small seed are worthy of note. Accurate measurements were made of the cotyledons of several kinds of seeds, and their surface areas determined with a planimeter. The average inner surface areas for several kinds of seed are given below stated in square inches.

SURFACE AREA OF LARGE AND SMALL SEED

  Large seed Small seed Average gain
for large seed
Cranberry bean 0.194 0.142 0.052
Garden pea 0.177 0.130 0.047
Hubbard squash 0.225 0.156 0.069
Horticultural bean 0.27 0.2 0.07
Sweet pea 0.063 0.037 0.026
Radish 0.018 0.016 0.002

Although the differences are not large, when expressed as percentages they range from 13 percent to 70 percent. They serve to explain in part the contrasts in seed weights and to throw light upon the subsequent growth phenomena. Initial differences of embryonic plants may be of as much importance in the initial growth of plants as an extra amount of food, which some writers in discussing farm seeds assume to be the point of advantage. Undoubtedly the extra food stored about the embryo, as in case of radish, or within the embryo, as in case of bean and squash, is an important factor; but in view of the structural differences of embryos it can hardly be assumed to be the only factor. However, wholly aside from considerations of the food supply, of the size of the embryo, or of the weight of the seed, there is likely to be something else to be reckoned upon in plant growth which cannot be measured, an inherent characteristic which we may call constitutional vigor.

The differences in size of embryo parts are correlated with differences in weight. Large seeds weigh more than small seeds. The greater surface area of the cotyledons has been noted. Are there differences in the weights of embryo structures? If so to what extent do seed coats and minor structures contribute to the total weight? In ascertaining the relative weights of seed coats and embryo, 200 large and 200 small Hubbard squash seeds were used.

  Total weight,
grams
Weight of seed
coat, grams
Weight of
embryos, grams
Weight of seed
coat, percent
Weight of
embryos, percent
Large, 58.4 15.8 42.6 27 73
Small, 49.2 12.9 36.3 26 74
Large exceeds small, 9.2 3. 6.3 1  

An inspection of the above figures indicates that the increased weight of large seeds is distributed through both coats and embryos and that the differences are so slight as to he negligible.

ECONOMIC CONSIDERATIONS

Will it pay to sort seeds according to size? A satisfactory answer to this question can scarcely be given unless full weight is accorded to many factors which must be taken into consideration in its formulation.

In the first place the kind of plant must be noted. The seed of some species and varieties differ more than others, and, consequently. the relative numbers of large or of small seed vary. In cucurbitous plants, seed of extreme size are few in number, scarcely more than 15 percent, whereas with lettuce the large and small together may total 40 percent of the entire number. The lesson which this bulletin teaches is the discard of small seed and the use only of the large and the medium sizes. This plan entails some loss of seed, which would amount to but little, except with high priced seed, but would effect a considerable saving in the matter of production. Moreover, the price of seed is one of the smaller items in the cost of producing a crop. It is better to waste at the beginning of a season rather than at the end, to waste seed rather than the time and labor expended in growing a crop. It costs little more to nurture a good crop than a poor one. Why not insure the crop so far as seed selection can accomplish that end?

Again the answer to the question whether it will pay to sort and discard depends upon the importance of the crop. In home gardening where space is not too valuable and where successive rather than simultaneous harvest obtains, it will scarcely be found worth while. In commercial work, on the other hand, where intensive methods prevail, whether indoors or outside, and where earliness and uniformity are prime requisites, it should pay. In market gardening operations, lateness of maturity and lack of uniformity may destroy profits or even turn profit into loss. A multiplicity of plants too small or too poor to market, may be due to the use of small seed interspersed with larger seed. Since earliness, uniformity of harvest, size of crop and quality appear to be more apt to obtain when large seed are employed, and since the reverse of these conditions generally attends the use of small seed, when a "sower (goes) forth to sow" he will consider not only whether his seed fall "by the way side," "upon stony places," or "among thorns," but also what manner of seed he soweth. The practical man will judge in advance the importance of his crop in relation to this matter and be governed accordingly.

Finally, the use of small seed is likely to result in many missing plants. The germination percentage of small seed is apt to be low. Many seeds fail utterly to germinate and thus contribute to unevenness of stand; and low yields are made yet more low because of a too scanty plant population. Such skips are much more frequent with small than with large seeds. Their use may therefore become a limiting factor in plant production.

In this connection the importance of good heritage must not be overlooked. Size is only one of many factors which influence yield. Parentage also functions. The immediately practical point to apprehend may be thus categorically stated: First, select seeds from good plants. Determine the attributes most worth while and which it is desired to perpetuate. After the first selection has been made based on general merits as well as on special characteristics, then use only the larger sized seed, discarding all inferior sized ones, even though their ancestry be good. Good seed is best secured by this double system of selection.

As a general proposition it will pay to discard small seed. A set of sieves for the different sized seed can be provided at small expense, and the regular use of this device will enable one quickly to eliminate seed of inferior size. Such mechanical means, however, are not accurate enough for experimental work. Sieves sort and grade only according to size and not with respect to weight or in relation to quality. Hand sorting is the only method that is reliable under such circumstances, for then account can be taken of wrinkled seeds, of shrunken specimens and of seed that is otherwise unfit or incapable of high yields. In the case of very small seed, like lettuce and radish, it is likely that a combination of the mechanical and hand methods will be found both desirable and necessary. The sieve should be used first, as an aid to and substitute for most of the hand sorting, and then hand work may be resorted to in order the further to eliminate any seed that are unfit for use.

SUMMARY

The trials reported upon herewith show that a distinct advantage follows the use of heavy or large seed. The results attained with seed sorting with sundry plants are of assistance in the explanation of several points relating to the differences in maturity, lack of uniformity in size, season, and quality of production.

The advantages which accrue from the use of large and heavy sweet pea seed are earlier blossoming, a larger total number of blossoms, and a larger number of blossoms of good quality as indicated by the size of bloom and length of stem. Plants grown from large seed are heavier and bear more and longer lateral branches.

Hubbard squash and sweet pumpkin respond well to seed selection. Plants grown from small seed yielded a larger number and a greater total weight of fruit, but were, however, markedly inferior with respect to number and weight of ripe fruit. Special precautions were found to be necessary in order to avoid a mixed heritage of seed.

The merits of large seed in lettuce culture are shown in the production of larger seedlings, an increased weight of edibly matured plants, which displayed better heading-up capabilities, earliness and uniformity in filling the heads; in short, augmented earliness and quality.

Heavy spinach seed outclasses light seed in weight of plants, width and number of leaves, height of main stalk and general earliness.

Parsley, a biennial herb which quickly renews its top when cut back, gives not merely larger and earlier foliage, but showed greater recuperative powers when the larger seed is used.

Radishes, one of the shortest of short term crops, show good gains in favor of large seeds. Sorting the sizes from the same parent gives as great a contrast as sorting composite samples, showing that gains may be made irrespective of the nature of the parent plants. Large seeds give a more uniform crop ready for use about one week in advance of small seed.

Trials with beans resulted in favor of the large seeds. The advantage accruing from the earliness of the product grown from the large seed is somewhat offset by its later germination.

In a single trial garden peas made little or no response to size selection of seed, although a slight gain was recorded for plants from large seed if allowed to mature; but no gain was observed when harvested as green peas.

The weights and sizes of plants compared at different stages of growth show that a continuous and permanent advantage exists in favor of large seed. Plants grown from large seed possess more leaves of greater surface area and hence have greater assimilative powers.

The place origin of small seeds in pods of different plants show a chaotic distribution. In beans, 49 percent were found to occur in the basal end, while 18 percent occur in the middle of the pod. Garden peas follow a more regular distribution, for the small seeds are almost always found at the ends of the pods, with one end as prolific as the other.