196 



fact, nothing but individual guessing from personal observation or des- 

 ultory facts. Now the difticulty is this : In estimating average rates 

 of yield per acre of any crop whatever, farmers almost invariably as- 

 sume too high a figure ; whether from pride in the reputation of their 

 neighborhood for fertility and good culture, or from the controlling im- 

 pression made upon their minds by the appearance of the best fields. 

 On the contrary, with reference to comparative aggregates ot yield, to 

 an inquiry whether the total area in a given crop is greater or less, or 

 the harvest better or worse than usual, their judgment is usually accu- 

 rate. If there is any bias, it will be likely to be in the direction of an 

 underestimate. These general facts will be acknowledged by all who 

 have any practical knowledge of crop-statistics ; they have been eftect- 

 ually attested by the experience of the past ten years of this Depart- 

 ment. This being the case, all our estimates of yield per acre are now 

 made, not alone from the direct estimate of yield of counties of all 

 crops, but after harmonizing resujts of reported local changes in the 

 rate of yield and in comparative aggregate of production. 



With regard to cotton, we have obtained an actual census, wherever 

 it was possible, for a given district, whether a large or a small portion 

 of a county, both of bales produced and the number of acres upon 

 which they were grown. Our correspondents were urged to avoid esti- 

 mates, and give only ascertained facts. So a census of a few farms 

 would be taken here, a neighborhood there, and occasionally a larger 

 district. The local officers were in some instances enlisted in the work. 

 The results are not complete, and are not deemed quite sufficient for an 

 authoritative estimate which can be relied on, in the future, as a per- 

 fectly accurate basis of comparison ; yet they are too important to be 

 withheld, and will be received by a fair-minded public as the best known 

 data for an approximate estimate of the real area in cotton. 



In the following statement the number of counties are given in which 

 such enumerations were made, either for their total or partial area, 

 respectivelv, with the number of bales produced on the acreage given, 

 in 1874 : 



States. 



Coun- 

 ! ties. 



North Carolina. 

 South Carolina. 



Georgia 



Florida 



Alabama 



MiSHigsippi 



Louisiana 



Texas 



Arkansas 



Tennessee 



Total 811,198 



70, 402 

 84, 249 



199, 810 

 17, 513 

 62. 534 



lO.-i, 280 

 56,450 



139, 762 

 49, 544 

 25,654 



Acres to 

 ithebale. 



202, 

 271, 

 707, 



67, 

 245, 

 308, 

 124, 

 404, 

 140, 



83, 



412 

 477 

 187 

 751 

 094 

 060 

 016 

 005 

 690 

 934 



2.87 

 3.21 

 3.54 

 3.86 

 3.92 

 2.92 

 2.19 

 2.89 

 2.84 

 3.27 



3.15 



This makes the average in over 40 per cent, of the cotton-breadth of 

 Georgia, 3.54 acres per bale, while the official report of the State of 

 Georgia makes the number 3^ acres — results almost exactly alike, as in 

 the reports of the area of the present crop. 



Without perfecting these estimates, it is evident that the true acreage 

 of 1874 cannot fall very much short of three times as many acres as 

 were produced of bales in 1874, or at least 11,000,000 acres, possibly a 

 figure a little larger. This corresponds with the mass of facts brought 

 to our observation annually for two years past. 



