102 



Housing and Feeding of Poultry. [may, 



matically falls as that already in the tray is consumed, 

 and the birds can eat whenever they feel inclined. Such a 

 system saves considerably in the labour of feeding. But a 

 further point is that meal only is given in this way. Hitherto 

 the majority of poultry keepers have believed that to keep 

 food constantly before the birds would be wasteful in the 

 extreme, and that the true method is to give just as much food 

 as will be eaten readily twice a day. Further, that whilst 

 grain is given in dry condition, meal must always be prepared 

 by mixing with water into a dry, crumbling mass. Hence " dry 

 mash " feeding is against all the canons hitherto accepted. 



With a view to testing these methods, a series of experi- 

 ments have been conducted at the College Poultry Farm, 

 Theale, in connection with University College, Reading, during 

 the last twelve months, of which complete records are now 

 given. 



Colony System of Housing. 



Location. — For the purpose of this experiment a section of 

 one field in permanent pasture has been allocated. It consists 

 of 3,154 square yards, or rather less than two-thirds of an acre. 

 It is good land with a capital belt of trees and hedgerow on 

 the south and east sides, where is a running stream, so that 

 the grass very seldom dries up, even in a hot summer. The 

 shape is uneven, almost triangular. A fence of 6 ft. wire 

 netting encloses it completely, and the birds are not allowed 

 out on the adjoining fields. The grass kept abundant 

 during the entire year, so that there was always a plentiful 

 supply of green food, but, it may be noted, the summer of 

 1907 was cool and moist, and there was no check to. the growth 

 of herbage. Within the enclosure are two ordinary open- 

 fronted poultry houses, but without scratching sheds. There 

 are no divisions of the ground, over which all the birds can 

 roam as they think fit. 



Birds used. — Fifty Buff Orpingtons were selected for this 

 experiment, namely, forty-eight pullets and two cockerels, all 

 of which were hatched in the month of May, 1906, from 

 purchased eggs. As records were kept from January 16th, 

 1907, they were at thai; time about eight months old. 

 Probably a much better result would have been obtained had 

 earlier hatched pullets been chosen, and from selected stock, 



