No. 1..

No. 2..

No. 3.

No. 4.

No. 5.

No. 6.

No. 7.

high and had 4 to 6 leaves, were spaced 16 inches apart. The singlestalk rows were thinned on June 24, when the plants, 8 to 10 inches high with 10 to 12 leaves, were spaced 6 to 8 inches apart.

Only one picking was made. The yield of each row as reported by Mr. Peterson is given in Table XX.

TABLE XX.-Row yields obtained in a single-stalk culture experiment with cotton conducted in 1915 by B. C. Peterson, Vanceboro, N. C.

Table XX shows that in 7 of the 10 instances single-stalk rows yielded more than the old-method rows. The difference in total yield, however, due largely to a poor stand, was only 18 pounds, or 3 per cent, in favor of single-stalk culture.

TABULAR SUMMARY.

The yields of seed cotton obtained from the 21 experiments discussed in the preceding pages are summarized in Table XXI.

TABLE XXI.-Summary of yields obtained in 21 single-stalk culture experiments with cotton conducted in 1915 in 9 parishes and counties in Louisiana, Arkansas, and North Carolina.

Single stalk.

Old method

Increase (per cent):

Gain or loss (−) for single stalk

193 336 645 123 99115 79 564 1,074 582 699 1, 497 850 765

620 649

505 226 287 227
380 177 207 206
350 211 149
300 178 140

229 338 780 112 96 90 82

686 1,004 540 582 1,347 782 722

510 631

295 195
262, 160

100

120

39
31

353
78

324

325

362 274

169
134

47

33

855 606 287 376 488 531 645 123 199 115 118... 917 1,074 582 699 1, 821 850 765 1, 029649 680489,207,346 491,498 780 112 216 90 113 764 1,004 540 582 1, 672 782 722

817 631

26 24 39 9-0.6 7-17 10-8 28 4 11 20 7 8 20

9 9 6

26 3

1 Actual yields not reported.

In six of the nine experiments conducted in Louisiana the total yield of seed cotton favored single-stalk culture by 9 to 39 per cent. In the other three experiments the total yield favored the old method by 0.6, 8, and 17 per cent, respectively. In the latter instance only one picking was made, and this so late that it is feared some cotton was lost, particularly from the single-stalk rows, which probably opened earlier than the others; also the reports of this experiment are so defective as to throw some doubt upon their record value. In the second instance the experiment was not inspected by the writer and only the yield records are available, so that it can not be stated with assurance that single-stalk culture was properly applied. In the first instance the stand was so poor that the new system could not be applied to good advantage, and it appears, moreover, that the first picking was made too late to secure the full yield. More than one picking was made from six of these experiments. In four of the six instances single-stalk culture yielded 3 to 32.9 per cent more seed cotton at the first picking. The yields from the other two experiments were influenced by thinning that was too late and by a poor stand, respectively. In five of six instances single-stalk culture yielded 6 to 22 per cent more at the second picking. A third picking was made in only one instance, and this favored singlestalk culture by 26.1 per cent.

There were three experiments in Arkansas, and in total yield these favored single-stalk culture by 4, 11, and 28 per cent, respectively. In one instance, where thinning was done a little too late, the old method yielded 4 per cent more seed cotton at the first picking, but this was offset by a 26 per cent increase for single-stalk culture at the second picking.

In all of the nine experiments in North Carolina, single-stalk culture gave greater total yields than the older methods, the differences varying from 3 to 26 per cent. In one instance, where the time of thinning was too late, the old method gave 22 per cent more seed cotton at the first picking, but this was offset by a 353 per cent increase for single-stalk culture at the second picking. In another instance, where three pickings were made, single-stalk culture yielded more seed cotton at each picking than the old method, the difference for each picking being 22, 32, and 42 per cent, respectively.

YIELD AND QUALITY OF LINT.

Having seen the effectiveness of single-stalk culture in increasing the yield of seed cotton over that of the usual systems of culture, it is of importance to know whether the lint was affected by the new system. It might be expected that the suppression of vegetative branches and the closer spacing of the resultant smaller plants would so affect the yield or quality of lint as largely to offset the advantage

gained through increased yield of seed cotton. This condition does. not exist, however, as shown by the results of determinations made in connection with the present experiments and those previously reported.

LINT PERCENTAGES.

Table XXII gives the per cent of lint as determined in samples of seed cotton sent in by five farmers from their respective experiments. Reference to Table I, page 4, will show the locations of these farmers, which fairly represent the general distribution of the experiments. One is in south-central Louisiana, one in northeastern Louisiana, one in southwestern Arkansas near the ArkansasLouisiana line, and two are in North Carolina. The varieties used in these experiments were, respectively, Half-and-Half, Sugar Loaf, Triumph, Prolific (probably King), and Cleveland Big Boll.

TABLE XXII.-Percentage of lint in samples of seed cotton taken in five single-stalk culture experiments with cotton conducted in 1915 in Louisiana, Arkansas, and North Carolina.

Table XXII shows practically no difference in the average percentage of lint obtained under the different systems of culture. The difference is remarkably small in view of the great differences in percentage among the individual samples from some of the experiments. The greatest difference in average percentage of lint is 1.2, and this favors single-stalk culture. In three other instances the differences range from 0.6 to 0.8 per cent in favor of the old method. In the remaining instance the average percentages are identical.

RELATIVE ABUNDANCE OF LINT.

The percentage of lint in itself is no more a safe basis for comparing the effect of different cultural methods than it is for judging the relative value of varieties.'. The percentage of lint would vary

1 Cook, O. F. Danger in judging cotton varieties by lint percentages. U. S. Dept. Agr., Bur. Plant Indus. Cir. 11, 16 p. 1908.

Meloy, G. S. Lint percentages and lint index of cotton and methods of determination. U. S. Dept. of Agr. Bul. 644, 12 p., 2 fig. 1918.

materially with the size of the seed if the abundance of lint on the surface of the seed remained constant. Conversely, if the abundance of lint were not constant, the lint percentages might appear uniform even though the size of seed varied considerably. Thus it is desirable to know whether the size of seed and the amount of lint per seed were affected by single-stalk culture. The weight in grams of 100 seeds taken from each of the samples discussed in connection with Table XXII is shown in Table XXIII, together with the lint indexes, or the grams of lint on 100 seeds, of the same samples.

TABLE XXIII.-Weights of 100 seeds and lint indexes (grams of lint on 100 seeds) of five different varieties of cotton as determined from samples of seed cotton grown in five single-stalk culture experiments with cotton conducted in Louisiana, Arkansas, and North Carolina.

It will be seen from Table XXIII that, while there was considerable variation in the weights of seed representing either of the systems of culture, the average weight of 100 seeds was about the same for each system. This fact, in addition to the fact that there was no significant difference in the percentage of lint, would indicate that the density of lint on the seeds was about the same under the different systems of culture. That is, we would expect to find that the weight of lint per seed did not vary significantly. The figures in the second part of Table XXIII show that such was the case. The lint index, or the number of grams of lint on 100 seeds, is seen to vary only slightly and in direct proportion to the size of seed.

GRADE AND LENGTH OF LINT.

Samples of lint in each of the experiments discussed in connection with Tables XXII and XXIII were submitted to Mr. Fred Taylor, cotton technologist of the Bureau of Markets, for a report on the grade and length of the lint produced by the different systems of culture. Mr. Taylor's report is embodied in Table XXIV.