278 TOBACCO PRODUCTION IN THE UNITED STATES. 



Nessler (op. tit., p. 138), to improve the combustibility of tobacco, recommends that it should be moistened with 

 a solution of acetate, or better, carbonate of potassium, adding " the best result was always obtained with an 

 aqncoiiK solution of potassium carbonate". 



While it is quite conceivable that a marked improvement should be attained by adding potassium acetate to 

 tobacco containing an excess of citric or malic acids, inasmuch as these acids would displace the acetic acid, it is 

 improbable that any good result should accrue from the addition of alkaline citrates or malates on the general 

 principle of increasing the percentage of organic salts in the tobacco, inasmuch as it is evident from the foregoing 

 results that an increase in the absolute amount, of citric and malic acids has a disadvantageous effect on the 

 combustibility of the tobacco. 



ANALYSES OF TOBACCO SOILS. 



These analyses of soils were made on samples from : First, the Housatonic valley, at New Milford, Connecticut; 

 second, the Clarksville district, Montgomery county, Tennessee, which produces the best variety of " shipping" 

 or " export" tobacco ; and third, Granville, North Carolina, where the best variety of the North Carolina Lemon 

 Yellow tobacco is grown. The two samples last named were furnished by Colonel Killebrew, and, according to him, 

 are fairly representative specimens of the best varieties of these respective soils. The first sample was taken by myself 

 at New Milford, from a field which, according to my informant (J\lr. Isaac B. Bristol), produces the best quality of 

 Housatonic Valley Seed-Leaf tobacco. 



The general character of the samples was as follows : 



No. 1. New Milford, Connecticut. Hill soil. Rich, dark loam, somewhat sandy in character. Average depth, 

 about 12 inches, with loamy subsoil extending to the bed-rock (gneiss). Geological character of the surrounding 

 rocks: Granite, gneiss, and mica, and hornblende slates. The rock fragments found in the sample were chiefly 

 granite, quartzite, mica slate, and, more rarely, hornblende slate. The field had been under cultivation six years, 

 exclusively on tobacco. Average yield, 1,500 to 1,800 pounds (dry weight) per acre of tobacco of fine quality. 

 Manured in the spring with 20 cart-loads of barnyard and stable manure per acre. The sample was taken November 

 30, 1880. The last application of manure was in the spring of the same year. 



No. 2. Clarksville, Montgomery county, Tennessee. Virgin soil. Rich, dark, clayey loam. Average depth 5 

 to 6 inches. Incumbent upon a highly ferruginous clay, the latter being intercalated with beds of chert, varying 

 from 1 to 3 feet in thickness. According to Colonel Killebrew this soil belongs, geologically, to the Lithostrotian 

 bed of the Siliceous group of the Lower Carboniferous formation. The small quantity of rock fragments contained 

 in the sample consisted chiefly of highly decomposed feldspar. The land adjoining the woods where this sample 

 was taken, produced, when first opened, 1,500 pounds of most excellent tobacco per acre, and has continued to- 

 produce it. in rotation with wheat and clover, for twenty-two years, with but little diminution in fertility. The 

 Clarksvilli; tobacco, marked No. 7 on the schedules of tobacco analyses (Tables I to III), is, according to Colonel 

 Killebrew, fairly illustrative of the character of that produced upon this soil. 



No. 3. Qranville, North Carolina. Light, gray sand, with porous cream-colored subsoil. According to Colonel 

 Killebrew this soil belongs to the Laurentine series of the Archrcan age, and is derived from granitoid rocks and 

 quartzites. Numerous trap or intrusive rocks are found in the vicinity, but the soil arising from the disintegration 

 of these will not produce the yellow tobacco, nor does this tobacco attain a high degree of excellence on soils 

 containing much ferric oxide. When the subsoil is red or clayey, heavy tobacco will grow, but not the yellow 

 leaf. The rock fragments found in the sample consisted chiefly of quartz. The field from which the sample was- 

 taken had been used for tobacco for six years in succession, but was previously an "old field" that had been 

 exhausted by long cultivation and allowed to lie uutilled for some 14 years previous to being used for tobacco, 

 and had become covered with "old-field pines" (Pimm tceda, Michaux), persimmon bushes, etc. Concerning the 

 manuring to which this field was subjected, I quote the following from Colonel Killebrew: "The soil is simply a 

 sponge, intto which just enough fertilizing matter is put to bring the plants to the proper size, and no more." 







METHODS OF ANALYSIS. 



1. MECHANICAL ANALYSIS. In the mechanical analysis of soils, which operation has for its object the 

 separation of the particles of the soil according to their size and the determination of the proportional quantity of 

 the particles of different sizes, the following method was adopted, viz: 



The whole sample was first passed through a sieve of perforated metal having holes of 3 millimeters in diameter. 

 The weight of the particles remaining on the sieve was then determined and also that of the portion passing 

 through the sieve ("fine earth"). The last-named portion constituted the material for all of the subsequent 

 operations of mechanical and chemical analysis. 



Thirty grammes of the "fine earth" were boiled out repeatedly with water, as recommended by Wolff (Chcm. 

 Vntersuchung landwirthnchaftliclier Stoffe: Berlin, Wiegant & Hempel [3tc Aufl.], 1875, p. 4), until the lumps were 

 disintegrated and the clayey portions separated from the sand. The material was then successively washed through 



perforated metal sieves, the holes in which were respectively l mm , 0.5 mm , and 0.25 mm in diameter. The portions 

 era 



