Fumigation with D D mixture has increased the chlorine content of the leaf and at times greatly lowered fire-holding capacity. In extreme instances, 10 gallons an acre of the material in the row 3 weeks before planting have raised the chlorine content of the leaf a full 0.60 percent higher than unfumigated tobacco. Calculated from leaf samples analyzed and assuming normal yields, that amounts to approximately 14 percent of the original 60 pounds of chlorine applied as fumigant. At other times the chlorine increase is scarcely perceptible and is not enough to affect fire-holding capacity.

The tolerance limit for chlorine in tobacco varies with the type. Some flue-cured tobaccos may tolerate up to 1 percent of chlorine as a maximum, if other qualities are suitable, but half that amount is excessive for cigar types. Low chlorine is always preferred.

With both fumigants, the greatest increase in bromine or chlorine and the lowest fire-holding capacity followed row treatments of 10 or more gallons an acre 3 weeks before planting. Packing the beds when it was not necessary to do so had similar adverse influences on leaf composition.

It might be assumed that such unfavorable effects on quality caused by excess chlorine and bromine would be reflected in lower grades and hence in lower prices for the tobacco grown on fumigated soil. That perhaps unfortunately has not been the case. Actually tobacco with an excess of of chlorine or bromine, and, for that matter, tobacco low in sugar and high in nicotine, may grade and sell very well. Over a 6-year period at Tifton, tobacco from fumigated plots averaged 20 percent or less of undesirable grades (common leaf and nondescript), while leaf from untreated plots averaged 30 percent of those undesirable grades. Also tobacco from the fumigated plots has repeatedly sold for a higher price than leaf from diseased, untreated plots. In fact, in an extensive series of comparisons the over-all average has been 3 to 8 cents a pound in favor of leaf from fumigated plots. That price differential was caused primarily by visibly poor tobacco from the untreated diseased soil. Where both fumigated and untreated plots were free from disease damage, usually there was no difference observed between grades and prices.

It has been possible with fumigation to produce tobacco that graded poorly. That has happened when tobacco was grown on fumigated land that contained more than the usual amount of organic matter. Under those conditions, fumigation delayed nitrification and the result was excess nitrogen late in the season. As a result, the top leaves remained green; that in turn lowered leaf grade and adversely affected leaf composition. Again, when heavy applications of fumigants were made 2 weeks before planting and root injury occurred, the resulting cured leaf has tended to be soggy and to turn dark upon aging.

The best protection against undesirable quality effects is to use the minimum effective dosage of fumigant, treat well in advance of planting, and treat only land that does not remain waterlogged long after fumigation.

Quality effects with D D mixture are further complicated by the fact that some tobacco fertilizers contain up to 2 and 3 percent chlorine. Some Georgia soils produce tobacco containing up to 1 percent chlorine or more when the fertilizers are used. D D fumigation of such soils tends to make a bad situation worse, unless chlorine is eliminated from the fertilizer. With a non-chlorine fertilizer and with a 6-week interval between treating and planting, the chemical composition of flue-cured tobacco after D D usually has been satisfactory. In wet soils an interval of 4 weeks between treating and transplanting has resulted in chlorine injury even when chlorine was eliminated from the fertilizer. The unpredictable breakdown of both fumigants, particularly in wet soils, is added reason for using the smallest applications. Low quality resulting from delayed nitrification, particularly in the presence of excess organic nitrogen, is reason for using minimum rather than maximum rates of nitrogen fertilizer.

A combination of 100 to 150 pounds of granular calcium cyanamide plus 2 quarts of soil fumigant (40 percent ethylene dibromide or D-15 mixture) to 100 square yards is one of the most effective and practical plant-bed soil treatments for control of weeds and nematodes in southern areas. Less damping-off usually occurs after cyanamide than in soils where that weed killer is not used. Three-fourths of the cyanamide is uniformly broadcast and worked in thoroughly not more than 3 inches deep. The soil fumigant is applied at a rate of 20 gallons an acre or slightly more. One-row applicators for field use are satisfactory. The furrows may be spaced 10 inches apart, instead of 12, to increase the rate slightly. The bed is again leveled and the remaining cyanamide is broadcast. This is raked in several times only 1 inch deep.

It is important that both fumigant and cyanamide be applied in the fall, or at least 60 days before seeding. If fumigant is not put in until a month or less before seeding, roots may be injured and plants stunted. In addition the plants may be induced to bloom prematurely in the field as a result of this plant-bed injury. The injuries can be avoided by early treatment plus location of beds in well-drained soil.

J. G. GAINES, a pathologist, has investigated tobacco diseases in Georgia since 1929.

T. W. GRAHAM, a pathologist in the Department of Agriculture, has received degrees from the Universities of Louisiana and Minnesota. He has studied extensively the nematode diseases of tobacco.