1916 many trees, 36 to 78 per cent, according to the position from which the material was taken, showed no decrease in ring width following turpentining. Also (see Table 3) 26 to 64 per cent showed no decrease in the amount of summer wood formed, and the summer wood appears to be one of the most readily affected structural features. In Table 4, covering both 1916 and 1917, similar results were shown; but the decreases in 1917, the second year of the operation, are somewhat larger in the case of this type of chipping than they were in 1916. The size of the resin passages also decreased somewhat as time went on. On the average a greater number of resin passages per unit area was formed on this tract in 1917 than on the double tract. In both 1916 and 1917, however, the number on the narrow tract exceeded that on the standard.

TABLE 2.-Comparison of the annual rings formed in round timber (1915) and turpentined timber (1916).

1 The narrow specimens showed more resin centers in the chips than were apparent in the standard and double specimens.

NOTE. The increment borings were made on the tree at a distance of 2 to 3 inches from the face and at the same height as the last streak. The chips were obtained at the cutting of the last streak.

TABLE 3.-Comparison of the amounts of summer wood formed in the round timber (1915) and the turpentined timber (1916).

1 The narrow specimens showed more resin centers in the chips than were apparent in the standard and double specimens.

NOTE. The increment borings were made on the tree at a distance of 2 to 3 inches from the face and at the same height as the last streak. The chips were obtained at the cutting of the last streak.

TABLE 4.-Comparison of ring width, summer wood, and resin passage formation for 1915, 1916, and 1917.

1 In 1916 material it was noted that the narrow had more resin centers than the standard or double. Minus sign (—) indicates decrease in width; plus sign (+) indicates same width or increase in width. TABLE 5.-Comparison of yield data for 1916 and 1917.1

TABLE 6.-Comparative yields per crop (10,000 cups) with respect to height of face (inches), 1916 and 1917.1

1 Compiled from monthly field reports for 1916 and 1917. 2 Minus (-) indicates loss; plus (+) indicates gain.

3 The narrow and double areas had four standard streaks before the experiment started. The height of the four streaks averaged 2.75 inches (average from 25 measurements). With this allowance, the streaks on the narrow and double areas averaged 0.30 and 0.32 inch, respectively. The corrected height for the doublo faces is 22.20 inches and for the narrow faces 11.38 inches, which, in this latter case especially, would further improve its rating.

If,

The comparative yields obtained are given in Tables 5 and 6. It is apparent that the total yield from the 1916 chipping was highest in the case of the double tract and lowest in that of the narrow. however, the narrow is compared to the standard with reference to height of face chipped, it is apparent that the narrow shows a gain of almost 40 per cent in the first year of the operation. During 1917, the second year of chipping, the results were even more strongly emphasized. The double showed about the same increase in total yield over the standard, but the narrow made a better relative showing than the year before, and nearly equaled the standard in total yield. With reference to the height of face, i. e., amount of chipping surface used to obtain the yield, the narrow showed almost a 50 per cent gain over the standard in yield per inch of height of face chipped.

In the second year in all crops there was some decrease in total yield. The 1917 comparisons, using the total 1916 yields from a crop as 100 per cent or the criterion for judging the relative yield of that crop, showed that the greatest decrease occurred in the standard tract. (Table 5.)

DOUBLE CHIPPING.

The special feature of this method, used for two years on this area at Columbia, Miss., was that the streak was cut at four-day intervals instead of only once each seven days. This type of chipping was used (Pl. IV, figs. 3, 5, and 6) on about 3,000 faces on the same kind of timber as that in the standard experiment. Only as much wood as was cut in the standard chipping was removed by this double method, since the dimensions of the streak specified were one-half inch deep and one-fourth inch high, cut twice weekly. The depth in general tended to average slightly less in the double than in the standard. During 1916 the chipping was carried on with a "00" hack (Pl. IV, fig. 2) and a streak averaging 0.32 inch was obtained (Table 6, footnote). In 1917 a "puller" (Pl. IV, fig. 6) was used, and a more accurate narrow chipping or rather "pulling " was obtained (average 0.26 inch) as is indicated in Table 6. This was also more accurate chipping than was obtained in 1917 on the single narrow-chipped area (average height of streak 0.34 inch), where a hack was used. It is of considerable interest to note that with this narrow chipping the double showed a smaller relative reduction in the second-year yield of turpentine, when compared to that of its first-year yield, than was shown by the wider-chipped (one-half inch per streak) standard. This was true in spite of the fact that the vitality of the double-chipped timber had apparently suffered rather more severely from the process of turpentining than had the standard.

In figures 3 and 4 are given the monthly observations on the five trees selected from the double area for 1916 and 1917, respectively (different sets of five each year). The same reduction as in the case of the standard was noted in the number of resin passages per unit area of the 1916 ring, as was observed in material cut at the level of the 1917 chipping. The tendency for fewer resin passages to be present at the end of the 1917 season than in midsummer was also observed. In 1917 practically all five trees from the double area showed that their wood formation had suffered as a consequence of that method of turpentining, and that they had not been able to recover, as many of the narrow-area trees had, or even been able to hold their own during 1917, the second year of turpentining, as some of the standard-area trees appeared to have done.

The results from the examinations of the 50 specimens collected at the end of the season each year are given in Tables 2, 3, and 4. In

Tables 2 and 3 results are given from a set of chips obtained at the last streak cut and a set of increment borings made 2 to 3 inches to the side of the face at the height of the last streak. The borings, in general, showed less effect from the turpentining than the chips, indicating that the response to the wound was not as marked tangentially or circumferentially as it was vertically. The double and the standard both showed more reduction in wood formation than the narrow, as indicated by ring width. Judged by the borings alone, the double showed slightly greater reduction than the standard. The chips, on the other hand, showed more reduction in ring width in the standard than in the double. This tendency, however, did not hold for the

amount of summer wood present, which was exceptionally reduced in the specimens from the double tract. It would appear that the double chipping produced a special response which was manifest in the increased ring width shown by the chip specimens and in the sustained relatively high yield for the second year, which has been mentioned. It would seem, however, that this response was accompolished at the expense of summerwood production and of the tree's vitality in general, judging by such indications as these and by the frequent occurrence of "dry" faces in this crop. The double also produced fewer resin passages than the narrow. In 1917 (Table 4) the reduction in wood formation in the specimens collected from 50