the same field man who took the original record goes back to the 

 farmer concerned and gets the additional data. 



A survey record generally consists of estimates made by the 

 farmer as to the details of his business for the past year. Most 

 farmers have the details of their business sufficiently in mind to 

 give a satisfactory, record, but in some cases it requires great skill 

 on the part of the field man in reducing his questions to the terms in 

 which the farmer thinks in order to get trustworthy information. 



If a record is grossly inaccurate this fact is easily detected by the 

 system of checking outlined below. Even if it is only slightly 

 inaccurate this fact will usually show because of the numerous 

 checks in the records themselves. The following is an outline of 

 the work to be done in checking a record : 



1. Compare the office sheet with the field blanks to see if all the 

 transfers have been correctly made. 



2. See if the total acres of crops, pastures, woodland, etc.. equals 

 the total farm area. 



3. Examine the yields of the crops per acre to see if they are 

 reasonable when compared with the normal for the year and 

 locality. (See next paragraph.) 



4. Examine the sales to see if prices and amounts sold are rea- 

 sonable. If the price is unusual, there should be an explanation 

 on the field sheet. In general, any unusual item on a field blank 

 should be accompanied by a note of explanation, to show it is not 

 merely an error] 



. 5. Balancing the live stock. Certain inconsistencies in the live- 

 stock record may be detected as follows: 



Cows. If the number of cows on hand at the first inventory be 

 added to the purchases and the sum be subtracted from the sum 

 of the second inventory plus sales, plus number died, the differ- 

 ence will show the number of heifers added to the herd during the 

 year. Suppose this difference is six and that in the first inventory 

 only four heifers were shown and that one heifer was purchased. 

 This shows a mistake of at least one either in the number of cows 

 or in the number of heifers. The number of heifers that became 

 cows during the year might be zero in such a case, no matter how 

 many heifers were on hand the first of the year. The point is: If 

 the number of heifers added to the herd during the year, as deter- 

 mined from the number of cows, is greater than the total number 



