The soil in the field on which the tests were made is Hartford sandy loam, and is fairly uniform in texture, drainage and fertility. It was laid out in plots of 1/40 acre, each containing four rows. Only three rows of each plot however, were harvested for the test because the fourth row was on the border and feeding from plots on each side of it treated in different ways. The whole series of seven plots was in triplicate making a total of twenty-one plots. All fertilizers were applied broadcast at one application about one week before setting. Rows were three feet four inches apart and the plants eighteen inches apart in the row. The variety of tobacco was Havana seed and all plots were set on the same day with plants as nearly uniform as could be obtained. The tobacco was primed in 1922 and 1923 but stalk cut in 1924.

After consultation with many growers in the Connecticut Valley and a careful study of previous field experiments, it was decided that a fertilizer containing approximately 260 lbs. ammonia; 225 lbs. phosphoric acid; and 240 lbs. potash to the acre would furnish ample quantities of these plant nutrients for an acre of tobacco. These amounts were considered as a basal ration in this experiment. Although such amounts of plant food materials are greatly in excess of that removed from the soil by the plant, farm practice in the Connecticut Valley has shown that the above quantities grow a good crop of tobacco. Accordingly, all the plots of this series. received approximately the same number of pounds of the above plant nutrients per acre.

The fertilizer treatment of the seven plots was as follows:
PLOT N1

BASAL RATION. 1-7 OF THE NITROGEN IN A MINERAL OR
INORGANIC FORM (SODIUM NITRATE). THE BALANCE IN
COTTONSEED MEAL AND CASTOR POMACE.

PLOT N2 ONE HALF NITROGEN IN INORGANIC CARRIERS (SODIUM NI

TRATE AND AMMONIUM SULFATE*). THE BALANCE OF
NITROGEN IN ORGANIC CARRIERS (COTTONSEED MEAL
AND CASTOR POMACE).

Carrier
Name

*These two materials used in amounts which theoretically would

not change the soil reaction.

PLOT N3

ALL NITROGEN IN MINERAL CARRIERS (SODIUM NITRATE
AND AMMONIUM SULFATE).

PLOT N4 ONE HALF NITROGEN IN MINERAL CARRIERS (POTASSIUM NITRATE AND AMMONIUM SULFATE*). THE BALANCE OF NITROGEN IN COTTONSEED MEAL AND CASTOR POMACE.

PLOT N5 ALL NITROGEN IN MINERAL CARRIERS (POTASSIUM NITRATE AND AMMONIUM SULFATE).

PLOT N6 ONE HALF NITROGEN FROM FISH, THE BALANCE FROM COTTONSEED MEAL AND SODIUM NITRATE (1-7 OF THE TOTAL NITROGEN IN THE NITRATE).

*These two materials used in amounts which theoretically would

not change the soil reaction.

PLOT N7 ONE HALF NITROGEN FROM FINE TANKAGE, THE BALANCE FROM COTTONSEED MEAL AND SODIUM NITRATE (1-7 THE TOTAL NITROGEN IN THE NITRATE).

From the above figures it is seen that the slight variation in the amounts of ammonia applied to the different plots is negligible. There is, however, a wide difference as to the type of substances used as nitrogen carriers. Plot N1 had one seventh of its nitrogen in mineral form; Plot N2, one half mineral nitrogen; Plot N3, all mineral nitrogen; Plot N4, one half mineral nitrogen; Plot N5,. all mineral nitrogen; Plot N6, and Plot N7, one seventh mineral nitrogen. The amounts of potash and phosphoric acid added to the above plots was the same in all cases, approximately 240 pounds K2O, and 225 pounds P2O, per acre.

SEASON OF 1922

The average yield and quality of the triplicate plots in 1922 as recorded by Chapman is presented in Table I.

TABLE I. EFFECTS OF DIFFERENT SOURCES OF NITROGEN on the YIELD AND QUALITY OF PRIMED HAVANA-1922.

Lbs. ammonia per acre Av. yield cured leaf, lbs. per acre

Plot

No.

Mineral

Organic

Genera!
Quality
Excellent
Fair
Poor

Fairly good
Excellent

Excellent

Good

A study of the above table shows that plots treated with all mineral nitrogen fertilizer had a tendency to give greater yields but poorer quality tobacco than plots receiving nitrogen of vegetable or animal origin. The quality of the tobacco on Plots N3 and N5, which received all the nitrogen in a mineral form, was so poor as to warrant an appraisal of fifteen cents a pound less than any other tobacco grown in this series. Not only was the tobacco of the first and second primings very poor, but the fourth was also of little character. The colors especially were poor, running almost entirely to a yellowish red, and not clean. Plots N1 and N6 were of the best quality and had a better finish than any of the

others. The body, grain, texture, and color, were highly satisfactory on these two plots, but plots N3 and N5 produced tobacco of harsh texture, poor color, and too close a grain. Plot N7, on which tankage was used, while yielding a large percentage of desirable colors, gave a more of less dark greenish tobacco after fermentation.

The growth of the tabacco on all of the plots of the nitrogen series was very satisfactory as far as judged by measurements. Topping time showed that there was less than ten percent difference in the average height of plants on the different plots. No fixed growth difference as to height could be attributed to the different fertilizer treatments. Although not suffering in growth, there was one particularly noticeable variation on these plots in 1922. Those plots which were fertilized with a large percentage of nitrogen in the mineral form developed a marked chlorosis, apparently directly proportionate to the amount of mineral nitrogen present in the fertilizer. Chlorosis was noticed especially at the base of the plant on those plots supplied with large amounts of mineral nitrogen. On those plots where half the nitrogen was supplied in mineral form and a half from organic sources, chlorotic effects extended upward on the bottom five leaves. Chlorosis extended half way up the plant on the plots which had all the nitrogen applied in mineral form; eight and sometimes ten leaves were affected. This chlorosis was in every respect similar to that called "sand drown" by Dr. W. W. Garner of the United States Department of Agriculture. Immediately adjacent to the nitrogen plots was another series of plots to determine the effects of magnesium deficiency. These plots were specially treated to show the cause of sand drown and were run in cooperation with the United States Department of Agriculture. The chlorosis in the nitrogen plots was in every way similar to the chlorosis which accompanied magnesium starvation. on the cooperative plots. For the season of 1922, at least, it was possible that the fertilizers high in mineral nitrogen did not supply adequate amounts of magnesium to fill the deficiency. The plots supplied with an abundance of organic nitrogen, on the other hand, did not develop chlorosis or "sand drown". Organic nitrogen carriers such as cottonseed meal apparently contain magnesium in amounts sufficient to meet the magnesium needs of the tobacco plant on an average Connecticut soil. The season of 1922 was very wet at times, and excessive leaching probably took place (which accentuated the effects). In a drier year, magnesium starvation probably would not follow the application of a heavy mineral nitrogen mixture. No chlorosis was noted in 1923.

SEASON OF 1923.

In 1923 the nitrogen plots were the same as to location and treatment as they were in 1922. It was a much more favorable season, inclined to be dry, while 1922 was extremely wet. The

tobacco was primed. Total weights of the triplicate plots were taken and figured to the acre basis as given in Table II. Each of the four primings was kept separate and a certain number of leaves taken from each for securing sorting data. All data (except total weights) are based on number of leaves. Burn tests were

also made. Sorting results are recorded in the table.

In general the results seem inconsistent and no general tendency can be traced. As in the preceding year, the use of all mineral ammoniates resulted in the highest yields. The data on burn and percentage of high grades do not show that the quality was worse than where only one-seventh of the ammonia was from mineral sources. The fact that the burn was always best on the first priming is very apparent. Both dry ground fish and tankage appear to be superior to the basal ration in yield, percentage of higher grades and burn.

SEASON OF 1924.

In 1924 the nitrogen plots were treated in the same manner as in 1923, being set with Havana seed on June 13 and 14, topped August 1 and 2, and harvested August 15 and 16.

Throughout the growing season no significant growth differences were observed between the different plots of the series. The tobacco on Plot N3 was slightly taller than that on the other plots but it also was more mature, being in full bloom when the others were only early and medium bloom at the date of topping. The average number of plants on the plots at harvest (about 160) indicates that the stand was practically uniform throughout the test.

The weather was extremely dry during both the growth and curing periods. The season was too dry for normal growth and a good cure. Scanty "damps" caused some difficulty in getting the crop in proper "case" until late in the fall. The tobacco was taken down in the middle of November and weights in the bundle were made November 28th, 1924.

In February, 1925, the tobacco was sorted by an experienced sorter, Mr. Frank Solkowski, and accurate weights were recorded on the different grades and lengths as indicated in tables III and IV.