and the Cattle Census. 413 



of March — and this difference of date will doubtless account for 

 the wide variation in the two results. 



This question of date is of the utmost importance in all statis- 

 tical inquiries, and above all in the case of stock, either live or 

 dead, which fluctuates so largely with the seasons of the year; it 

 is probably from this point being overlooked that "doctors " have 

 so much "disagreed" in their estimates of the cattle stock of this 

 and other countries. If all the counties of England and Scotland 

 were alike in their treatment of live stock, the difference of the 

 date of enumeration would be less important ; but as one county 

 breeds stock for the supply of another, which grazes or milks but 

 does not breed, the relative stock of these two counties will 

 depend entirely on the time of year at which their numbers are 

 taken. If this be not the explanation of such a fact, for instance, 

 as that Wiltshire is stated to have had 35,241 cattle, 418,197 

 sheep, and 27,998 pigs in 1854, whereas that county is returned 

 in 1866 as possessing 77,724 cattle (more than double the first 

 return), 596,822 sheep, and 61,012 pigs, it is difficult to under- 

 stand what is the cause of the difference. It can hardly be attri- 

 buted to any great increase in the proportion of returns sent in 

 this year, for in 1854 no less than 87^ per cent, of the schedules 

 were returned with the information supplied either by the occu- 

 pier or by the enumerator, and it is scarcely likely that a very 

 much higher proportion of farmers made the return on this last 

 occasion. Hampshire, Suffolk, Berks, Worcester, and Salop are 

 all credited with a great increase of cattle ; while in Norfolk, 

 the West Riding, and Brecknockshire — counties which showed 

 the highest proportion of complete schedules in 1854 — the cattle 

 appear by the Table to have decreased. The numbers of sheep 

 and pigs also differ so widely as to make it clear that the basis 

 of comparison is not uniform. I point out these divergences to 

 show how extremely difficult — nay, how impossible — it is to go 

 beyond the merest statement of fact in dealing with statistics 

 of so imperfect a nature as characterises all our agricultural data 

 up to the present time. Under these circumstances I have no 

 alternative but to let the figures of the recent cattle census speak 

 for themselves, until the Inland Revenue Department shall think 

 fit to enlighten us on those doubtful points which make their 

 return so unsatisfactory for purposes of comparative deduc- 

 tion. If that information be supplied it will convert the Return 

 of Live Stock in 1866 into a valuable point of departure for 

 comparison with subsequent enumerations, but without it the 

 thing will remain sui generis, an isolated fragment, useful to a 

 certain degree, but having no legitimate connection with any 

 future return. 



In the annexed Tables will be found all the particulars of the 

 Returns of Live Stock as laid before Parliament for each county 



