AGRICVLTIRAL AI'IMUHMtlATlON BII.I-, 11124. 419 



Hcrcujjc. wliicli arc Ix'iu^ put iiitn piju'licr which will luivr u vitv 

 important rfrcct on our ostiinatin^ w<»rk of tin- future. If vou arV 

 intcrt'sti'cl. 1 will nuMition thoni. One is hv jirttinj; ciinii);!! ukIivhIuhI 

 farmci's to rrpoit what tlu'v thnnsrlvcs )iavr ;^n»wu ami pr«*(lur<><|, 

 so that whni we <;rt a rjrcat many tliou»»an<ls of thi-sr n>jMtris put 

 to<;«*tlu'r w(> will liavc a fair samph* of tho whole rountry. Wi» havr 

 ma<le somi' stjidirs as to how many wr iummI to do that, and our 

 studies s(» far would indicate that if we «-ould ^et ahout 2 per cent «»f 

 the farmers of the country to tell us exactly what acrea;je thev have 

 in oro|)s we woidd have enou<;h t») jjive us a true pirturi' of what is 

 ha|)penin<r. We have done that this year in everv State, and our 

 revisions for Decemher will he hased larj^ely on those studies. Then 

 wo have encoin*a}4ed the developm«>tit of the assessor's census, u> you 

 tnif^ht call it, and in ahout 1') or 20 States the States have pH.s.s<><| 

 laws which r»'(|uire the assessors, at the tim«' of makini; their annual 

 assessments, to take n crti^us of the ncreML''' "fid tnindier^ of live 

 stock. 



In ahout four States the data have heccune so accurate that we feel 

 it is renllv hetter than th(> Inited States census so far as acreajje is 

 concerned. That is ««speciallv true in \Visc(»nsin ; in Minnesota this 

 year thev are taking it for the first time and it is extrenielv satis- 

 factory; Michigan has an annual assessors' census as well as Iowa, 

 Nohraska. Missouri, and Colorado. \\v are anxious to see and arc 

 encouraging the pa.ssage of State laws which will give a census in 

 every State. Then we have also, following out the suggestions made 

 by a statistical committee of experts who investigated our system 

 last winter, started a system of field counts, as we call it. that is. each 

 of our field statisticans takes a certain line of railroad or certain road 

 and he covers that same area each year. The arrangement now is 

 that he will take enough of those r»)a(ls in his State to give a goo<l 

 cross section of his State and will count the number of fields of each 

 kind of crop along those roads, makin<c an exact record every year of 

 what changes have taken place. In tne case of fields that are not of 

 good shape or ar(* of odd sizes we are trving the experiment of counting 

 the kind of crop that is opposite each telegraph pole. We find the 

 field count is a very satisfactory way of estimatmg the changes in 

 acreage from year to year in the East and South, but it is not so 

 satisfactory in the West. 



Some of our men are quite enthusiastic about the accuracy with 

 which changes in acreage can be determined by coimting the fields 

 from year to year over exactly the same area. In a State like (leorgia. 

 where it has been tried out to the greatest extent, it takes about 

 4.000 miles oi road, covered year after year in exactly the same 

 way^ to determine the change that has taken place during the year. 

 I might say that last year in South Carolina our statistician covered 

 about .3,000 miles of road and estimated the ct)tton crop base<l on 

 his survey. His estimate was about the same as that finally deter- 

 mined in December after the ginning returns were nearly all in. 

 All of the other sources of information gave a much lower acreage 

 than his survey, but as it was a new experinu'iit he wm'^ to tru<t the 

 results. 



