AREAS OF GREATEST PRODUCTION.-A statistical map accompanies this report which shows in each region the percentage of the total area planted in cotton; and since the differences in product per acre in the several agricultural regions are comparatively slight, this map shows also approximately the percentage of the whole crop produced in each region.

Upon examination of this map, or of the statistical tables given on page 60, we see that the central prairie region produces 40 per cent. of the entire cotton crop of the state, the oak, hickory, and long-leaf pine region 17 per cent., the Tennessee valley 12 per cent., the metamorphic 10 per cent., the Coosa valley 8 per cent., the gravelly hills 5 per cent., the short-leaf pine uplands 4 per cent., the Coal-measures region 3 per cent., and the long-leaf pine region less than 1 per cent. But these relations will be much more clearly shown if we take into consideration also the relative areas of these regions and rate them according to the number of bales to the square mile. The several regions will then rank as follows:

By this arrangement we are able to recognize three well-defined areas of large production in the state. These are: 1, the central cotton belt; 2, the Tennessee valley; 3, the Coosa valley. The first of these areas produces at least 60 per cent. of the cotton crop; the second, 12 per cent.; the third, 8 per cent.; while the remaining 20 per cent. is produced by the rest of the state.

The nucleus of the central cotton belt is composed of the 12 counties of the prairie region, together with Chambers and Lee counties and the southern portion of Tallapoosa county, of the metamorphic region. On each side of this nucleus there is a margin consisting of the adjacent portions of the counties of the short-leaf pine uplands and gravelly hills on the north, and of the oak, hickory, and long-leaf pine uplands on the south, in which the cotton production assumes nearly as great proportions as in the prairie belt itself. In this way are included the southern parts of Pickens, Tuscaloosa, Autauga, and Elmore counties, and the northern parts of Wilcox, Butler, Crenshaw, Pike, and Henry counties.

The width of this central cotton belt across the state is not much less than 75 miles, and the cotton production throughout the area thus defined is 20 bales and upward to the square mile.

Of the counties which form the nucleus of the central cotton belt Dallas has the largest total production, because of its greater area; but if we take into account the differences of area, Montgomery occupies the first place, producing 43 bales to the square mile. After Montgomery come Lowndes with 40 bales, Dallas and Bullock with 34 each, Chambers with 32, and so on down the list, ending with Tallapoosa, which has 17 bales to the square mile. It will be seen by reference to the tables that some of the marginal counties of this central belt show a yield for the whole county of 20 bales and upward to the square mile, which would, of course, bring up the yield of the portion of the county actually embraced in the central region.

The second well-marked area of large production is found in the Tennessee valley, with its greatest intensity along the immediate valley of the river, and the relative importance of the several counties here included in the production of cotton may be approximately measured by the relative proportion of the red or valley lands in the area of each. Applying the test of area devoted to cotton, we find that the counties rank as follows: Madison, Limestone, Lawrence, Lauderdale, Colbert, Jackson, Morgan, and Franklin; but in eliminating the disturbing influence of difference in area, and taking account also of the product per acre, their rank as regards the number of bales to the square mile becomes: Limestone, Madison, Lawrence, Colbert, Lauderdale, Morgan, Jackson, and Franklin. Of these counties only the first two produce over 20 bales to the square mile.

The third large cotton-producing area is the Coosa valley. In actual production to the square mile this falls behind the other two regions, being 13 bales, while that of the Tennessee valley is 15, and that of the central cotton belt 20 bales and upward. Both in total production and in cotton acreage, and especially in the number of bales to the acre, the counties of this region fall behind those of the two preceding. Talladega county stands first, with 17 bales to the square mile; then come Calhoun, Cherokee, Etowah, Saint Clair, and lastly Shelby, with 9 bales. We have thus seen where the great proportion of the cotton crop of Alabama, 80 per cent., is produced. As regards the rest of the state not much more need be said, except concerning the Coal Measures and the long-leaf pine regions, which show a production of 4 bales and 1 bale respectively to the square mile. With regard to the former region, it is to be remarked that the greater part of the cotton is produced in the valleys which traverse it, and not upon the soils of the true Coal Measures. Winston and Cullman may be taken as fair representatives of this region.

POPULATION AND COTTON PRODUCTION.-Taking the state as a whole, the cotton production of Alabama is 0.55 bale to the inhabitant, or little more than a bale for every two inhabitants. This proportion varies in the

different agricultural regions. In the Tennessee valley and the gravelly pine hills the proportion is exactly a bale to two inhabitants; in the short-leaf pine uplands the proportion is a little more than a bale to two inhabitants, and in the Coosa valley it is slightly less. In the central prairie region the proportion is a bale and a half, and in the oak, hickory, and long-leaf pine uplands a bale and a quarter to every two inhabitants. In the last-named region, however, there are parts of some of the counties immediately adjoining the prairie region in which the same proportion probably obtains as in the prairie region itself. The same is true of the counties of Chambers and Lee; so that for the great central cotton belt the proportion is about three-fourths of a bale to the inhabitant.

Following out these relations a step further, we find that over 55 per cent. of the colored population of the entire state is to be found in the central cotton belt, where about 60 per cent. of the cotton is produced. Something over 10 per cent. of the blacks are found in the second cotton area, the Tennessee valley, and about 5 per cent. in the Coosa valley. This accounts for more than 70 per cent. of the colored population, which is thus concentrated in the three large cotton-producing areas of the state, where about 80 per cent. of the cotton crop is produced.

The distribution of the whites in the same regions is as follows: In the central cotton belt, about 18 per cent; in the Tennessee valley, about 16 per cent.; and in the Coosa valley, about 12 per cent.; thus accounting for about 46 per cent. of the white population, as inhabiting the three large cotton-producing areas, against 70 per cent. of the blacks in the same areas. The other regions of the state which produce the remaining 20 per cent. of the cotton support 54 per cent. of the white population, but less than 30 per cent. of the blacks. Since the proportion of the white to the black population in the whole state is about 1.1 to 1, or not far from equal, it seems to follow that the greater part of the cotton crop of Alabama is produced by the negroes.

PRODUCT PER ACRE AND ITS RELATION TO POPULATION.-In product per acre Alabama stands No. 13 of the fourteen principal cotton-producing states of the Union. Other things being equal, the product or yield per acre may be taken as an index to the fertility of a soil, and if we apply this test to the several agricultural regions of Alabama they take the following rank:

1, Coosa valley; 2, Coal Measures; 3,oak, hickory, and short-leaf pine uplands; 4, Tennessee valley; 5, gravelly hills; 6, oak, hickory, and long-leaf pine uplands; 7, long-leaf pine region; 8, metamorphic; and 9, central prairie region.

Putting Chambers and Lee together with the counties which constitute the prairie region, we have the nucleus of the central cotton belt as above defined. In all these counties the average product per acre is 0.27 of a bale. This somewhat unexpected result cannot be considered as due to the relative infertility of the soils of this belt, for correspondents unite in giving as the average yield on the fresh lands of this region from 700 to 1,600 pounds of seed-cotton, or from one-half a bale to more than a bale to the acre, and the chemical analyses show that these soils in their virgin state are among the very best in the state. We are led, therefore, to the conclusion that the soils of the great cotton belt have been exhausted by improvident culture, and, as a matter of fact, we know that in many parts of this belt cotton is planted year after year upon the same soils without rotation with other crops, and without an attempt at maintaining the fertility by the use of manures. In the other parts of the state where cotton is produced a selection is generally made of the better soils, rotation of crops is more generally practiced, and in some sections fertilizers are in more general use.

That the character of the laborers and the system of farming practiced are largely concerned in determining the yield cannot, on general principles, be denied, and we find ample proof that these two things are responsible in no small degree for the results above shown.

The central cotton belt is generally a region of large farms or plantations, in which the laborers are chiefly negroes, as seen in the tables. As a rule, these laborers do not own the land, have no interest in it beyond getting a crop from a portion of it, which they rent either for a sum of money or for a share of the crop, and are not interested in keeping up the fertility, at least not to the extent of being led to make any attempts at the permanent improvement of the same. In the case of the owner of the land, while the conditions are different, the result is the same. He is, of course, interested in the improvement of his land; but to supply the fertilizers for a large plantation, when he cultivates it by hired labor, would cost more than he usually has to expend, and where the share system, or that of renting, prevails he is still further removed from personal care of the land; and thus from all causes there is an exhaustive cultivation of the land, without any attempt at maintenance or restoration of its lost fertility.

In addition to these, the system of advances or credit, so prevalent throughout the cotton-producing parts of the state, is not without its evil influence, for the laborer, and too often the owner of the land, is obliged to get advances of provisions from their merchants, for the payment of which the crop is mortgaged; and as cotton is the only crop which will always bring ready money, its planting is usually insisted on by the merchants making the advances and selected by the farmer as a means of providing for payment. In this way cotton comes to be the paramount crop, and there is little chance for rotation with other things.

In this connection it will be instructive to read the reports given under Part III, treating of cultural details. It will there be seen that the system of credits in the large cotton-producing regions prevails to such an extent that the whole cotton crop is usually mortgaged before it is gathered; and when we consider that the prices charged for provisions, etc., thus advanced are at least 50 per cent. higher than regular market rates, and that the cost of

producing cotton is given by our correspondents, almost without exception, at 8 cents a pound, it will need very little calculation to show that the laborer who makes a profit of only 2 or 3 cents a pound or $12 to $15 a bale on his cotton will have the chances too greatly against him ever to be out of debt to his merchants when he relies solely upon this crop to provide the money; and the exorbitant interest on the money advanced is not likely to be lessened so long as the merchants' risks continue to be as great as they are.

In the Tennessee and Coosa valleys, which are also large cotton-producing sections, a similar state of things may be observed. In Madison and Talladega counties the blacks outnumber the whites, and in both we find the product per acre falling far below the average of the region in which they are situated. Thus Madison shows a product of 0.28 bale, against the average of 0.32 for the whole Tennessee valley, and Talladega a product of only 0.36 bale, when the average for the Coosa valley region, of which it is a part, is 0.40. Wherever the black population is in excess of the white we may take it for granted that the system of large farms rented out to the negroes prevails, and the inevitable result of this system of farming thus becomes apparent in these sections also.

In the other agricultural regions of the state, and in most of the counties also of the Tennessee and Coosa valleys, the farms are, as a rule, small, and are cultivated by their owners, with the assistance of such labor as may be hired from time to time. In all these cases provisions are produced on the farm, and cotton is planted as a secondary crop. There is thus some chance for selection of the soils and for rotation of crops; and when a man cultivates his own farm fertilizers are in more general use, so that even with soils naturally much inferior to those of the main cotton-producing regions the average product per acre is much higher in these regions of small production.

In the Coal-Measures region, which takes rank as third in product per acre, there are no large farms, and the whites outnumber the blacks nearly 9 to 1 (a sure sign of poor soil), the farmers generally owning the land which they cultivate.

Until very recently only the lands of the valleys traversing the Coal Measures have been planted in cotton, so that the product per acre as given in the tables is an index rather of the fertility and capabilities, under proper culture, of the calcareous valley soils than of those directly derived from the rocks of the Coal Measures. In Winston and Cullman there are no valley soils, hence the product per acre of these counties may be taken as representing that of the soils of the Coal Measures generally, viz, 0.26 to 0.28. In these regions it is usual to plant only the better kinds of soils in cotton, and of late only with the application of some kinds of commercial fertilizers. These sandy lands, which have a clayey substratum, are more and more every year, with moderate quantities of fertilizers, coming into use in the production of cotton, and the same may also be said of the siliceous portions of the valley lands of the regions just spoken of. It is now thought to be pretty well established that these poorer sandy lands, with the aid of moderate quantities of fertilizers, make in the long run better-paying and more certain crops of cotton than the intrinsically better classes of soils without the fertilizers.

In the short-leaf pine upland counties the whites are more than twice as numerous as the blacks. The same conditions, therefore, hold here as in the case just mentioned. The product per acre is 0.33 per bale.

The soils of the gravelly hills are practically the same as those of the short-leaf pine lands, but the product is 0.30 bale to the acre, and that of the oak, hickory, and long-leaf pine uplands about 0.30, with likewise very similar soils. In these the two races are present in nearly equal proportions. These figures furnish an additional illustration of what has been shown above, viz, that the greater the proportion of blacks among the population the more prevalent will be the system of large farms worked on shares or by renting and the smaller will be the yield of the land so cultivated, because of the inherent vices of the system.

The concentration of the black population in the great farming regions of the state, which are also the regions of the originally most fertile soils, is amply shown by Table III; and so closely does this class of the population follow the best lands that the density of the colored population of any region might almost be taken as an index of the fertility of its soils. The white population is much more evenly distributed over good and poor lands alike, so that the proportion between the two races varies with the fertility of the soils. (a)

If we examine any county whose product per acre falls below the average of the region of which it forms a part, we shall find almost without exception that this is due either to the improvident culture which invariably attends the system of large farms (and the prevalence of this system is almost invariably shown by the preponderance of negroes among the population) or to the comparative infertility of the soil, as may be inferred from the preponderance of the whites. Thus in the county of Marion, in the short-leaf pine uplands, we find the product only 0.31, and we see that the county is inhabited almost entirely by white men. Again, in Pickens, there is a slight falling below the average, and here we find one-third more negroes than whites (large farms and bad culture). In the gravelly hills region Autauga falls below the average, notwithstanding it possesses some of the best second-bottom lands of the Alabama river. In this county the negroes outnumber the whites two to one.

a The negroes were originally brought together upon these great cotton-producing areas as the slaves of the wealthy planters who bought up the greater part of the best lands in the state. Since the war they have remained, practically speaking, in the same places where as laborers in the cotton-field (with which they were most familiar) they could always be sure of employment and of a good living without too severe labor. The social attractions also of these great centers of negro population have not been without their

In the oak, hickory, and long-leaf pine region we meet with some exceptions to the general rule. Most of these. however, admit of explanation. Thus, Choctaw follows the rule: excess of blacks, below the average of product per acre. Clarke, Monroe, and Wilcox form exceptions: large negro population and high product per acre. This finds its explanation in the fact that the cotton lands of those three counties are either the very best of river lands. such as form Black's bend, in Wilcox, or the equally fertile lime-hills or black-shell prairie lands, like those in the vicinity of Limestone and Flat creeks, in Monroe. Butler and Pike follow the rule: excess of white population, small farms, better cultivation, general use of fertilizers, all of which combine to bring up the product per acre above the average for the region. The lower counties, such as Conecuh, Coffee, Dale, and Henry, lie within the limits of the long-leaf pine region, where the poverty of the soil is more than an offset to the better cultivation practiced on small farms. Conecuh has also a large negro population on some of its best lime-lands to keep down the average product.

In the counties of the open piny woods there is comparatively little cotton produced, not much more than 1,000 bales in any of the counties. There is thus a selection of the best lands for the planting of cotton, and a correspondingly high yield in Baldwin, Washington, and Escambia, with the additional circumstance that in Washington there are fine lime-hills and shell prairies, with the highly productive river bottoms, which lie adjacent to the same, to bring up its average.

Covington and Geneva, in their very small percentage of blacks (only one in seven or eight), show that their low product per acre must be due to the original poverty of the soil, and the open piny woods which make so large a proportion of these counties would lead us to expect none but soils of much less than average fertility.

The position of Alabama as a state, next to the lowest in product per acre of the fourteen cotton-producing states, has already been justly explained by Dr. Hilgard as due to the exhaustion of the soils by bad or improvident culture, and to the fact that the system of returns to the soil is not yet in general practice, as is shown by the very limited use made of fertilizers. The conditions of the different regions as above set forth furnish ample illustration of the truth of this conclusion.

INFERENCES TO BE DRAWN FROM THESE COMPARISONS.-To recapitulate, the following conclusions seem, therefore, to be plainly taught by the discussion of the data contained in the tables presented on page 60:

1. That where the blacks are in excess of the whites there are the originally most fertile lands of the state. The natural advantages of the soils are, however, more than counterbalanced by the bad system prevailing in such sections, viz, large farms rented out in patches to laborers who are too poor and too much in debt to merchants to have any interest in keeping up the fertility of the soil, or rather the ability to keep it up, with the natural consequence of its rapid exhaustion and a product per acre on these, the best lands of the state, lower than that which is realized from the very poorest.

2. Where the two races are in nearly equal proportions, or where the whites are in only slight excess over the blacks, as is the case in all the sections where the soils are of average fertility, there is found the system of small farms, worked generally by the owners, a consequently better cultivation, a more general use of commercial fertilizers, a correspondingly high product per acre, and a partial maintenance of the fertility of the soils.

3. Where the whites are greatly in excess of the blacks (three to one and above), the soils are almost certain to be far below the average in fertility, and the product per acre is low from this cause, notwithstanding the redeeming influences of a comparatively rational system of cultivation.

4. The exceptions to these general rules are nearly always due to local causes, which are not far to seek, and which afford generally a satisfactory explanation of the discrepancies.

FERTILIZERS.

THE USE OF FERTILIZERS IN COTTON PLANTING IN ALABAMA.-In the foregoing remarks on cotton culture incidental mention has been made of the use (or, to speak more correctly, of the non-use) of commercial fertilizers, from which it may be inferred that systematic efforts at the maintenance of the fertility of the soils in Alabama are not generally made by the farmers. There is, however, probably not a farm in the state where the barnyard manure and composts produced on the farm are not spread upon the land. This kind of manuring is almost universally practiced, but in this way only a very small proportion of the land receives any assistance. In many sections cotton-seed, either alone or composted with other things, and in certain cases also the cottonseed-meal, are beginning to be somewhat generally used, but always as yet sparingly, and upon a small portion only of the whole area in cultivation. In the regions of small farms, and especially where the soils are originally not very strong, the use of commercial fertilizers, guanos, superphosphates, etc., is gradually extending, and more rapidly in the eastern than in the western half of the state. There are many soils until recently thought to be too unproductive for cotton planting which are now quite extensively used for this purpose, since it has been found that, with the use of small quantities of commercial fertilizers, better returns of cotton are realized from such soils than from better soils without the fertilizers. Instances of this are seen in the gray flinty lands of the Coosa valley, the sandy lands of the Coal Measures, and the Barrens of the Tennessee valley.

In the southern counties of the oak, hickory, and long-leaf pine uplands, and in the long-leaf pine region itself, the poor quality of the soils has compelled the farmers to use some means of bringing up the yield, so that the