640 iiGRICULTTJBAL STATISTICS. [OcT., 



give a result below or above the true figures. This will happen 

 whether the return is voluntary or compulsory. To take a 

 familiar instance. The Census returns are compulsory, but 

 it is well-known that certain female age-groups are definitely 

 inaccurate by reason of a general tendency to under-state the 

 ^ge. Any financial interest provides a still stronger motive. 



In stating the acreage under each crop and the number of 

 each class of live stock, the individual has no obvious reason 

 for deliberate misstatement, and it is on the whole easier to 

 give the true figures than to invent others. As the total 

 acreage of the farm is taken from the rate-book and the sum 

 of all the crops must make that total, a certain amount of 

 trouble, for which there is no sufficient motive, is involved 

 in compiling a statement which is untrue in detail. 



While, therefore, I agree that compulsory returns are on 

 ihe whole more accurate than voluntary returns, as well as 

 more complete, the effect on the general total is practically 

 :negligible in a comparison of figures collected under both 

 systems. 



Returns of production of crops were first collected in 1885. In 

 the nature of the case they cannot be obtained in June, but must 

 be collected after the crops are gathered. They are, in fact, 

 obtained in mid-October for the corn, pulse and hay crops, and 

 in mid-Xovember for potatoes and roots. The collecting officers 

 are supplied with schedules giving for each parish the acreage 

 of each crop as returned in June, and their duty is to enter 

 the estimated average production per acre of each of the crops. 

 They are instructed to arrive at their estimate after enquiries 

 of growers, thrashing machine owners, valuers, &c. Much, how- 

 ever, obviously depends on the judgment and experience of the 

 collecting officer. 



It is sometimes objected that estimates are not ascertained 

 facts — whic-h is undoubtedly true . The system ma v be sufficiently 

 defended on the gimmd that no other is practicable, but, in fact, 

 it may also be claimed that no other is superior. Two other 

 methods are possible. One is to obtain a return from every 

 gi'ower, and the other to rely on the ascertained results of a 

 number of sample plots. A return from every grower would, 

 of course, involve more trouble and expense and greater delay, 

 l)ut apart from this there are objections. In many cases the 

 grower has no a<?curate record of the quantity of the crops — 

 especially such crops as hay and roots which ai'e consumed on 

 the holding — but assuming he has the exact figures the difficulty 



