QUANTITY OF FOOD EATEN. 17 



foundation. A few years ago, ho-wever (in 1873 and 1874), 

 Mr. Harrison, Secretary of the London Amateur Pigeon Society, 

 made very careful actual experiments to solve the point, and 

 the results are sufficiently important and interesting to be put 

 on permanent record. In 1873 account was kept for thirty- 

 eight weeks, and when the figures came to be worked out, it was 

 found that each pair of birds, on an average, consumed one pound 

 thirteen ounces, or rather over a pint, per week. The pigeons 

 were Homing birds, and the grain used in this experiment con- 

 sisted solely of tares, purchased at 17s. per sack, thus bringing 

 the cost per pair per week to little over one penny. Tares are 

 scarcely ever so cheap as this, of course. In 1874 another 

 experiment was made for twenty-eight weeks on the same food, 

 with a result of 1^ pints per pair per week. Later experi- 

 ments made with grain at higher prices showed that, at 49s. 

 per quarter, the pigeons cost about 5s. per pair per annum, and 

 at 60s. per quarter, about 6s. per pair per annum, or say 1 Jd. 

 per week. It must, however, be remembered, on the one hand, 

 that Homing pigeons are generally fed at stated times, and then 

 by allowance ; and on the other, that few pigeons eat tares so 

 readily as other grain : at least, our own experience has in- 

 variably been that tares, however good, are left to the last. We 

 never kept any detailed account ; but our general experience 

 has been that pigeons kept confined cost us a fraction over 2id. 

 per pair per week, though some of this may have been owing 

 to the sparrows before mentioned. 



We strongly advise that the birds be "fed from a hopper, 

 and have the food always by them ; except where — as with 

 Flying Tumblers and Homing birds — stated feeding times bear 

 an important part in their training. It does not very much 

 matter how the hopper is constructed, so long as it is impos- 

 sible for the birds to stand upon or foul the feeding apertvires. 

 Fig. 8 shows Mr. Fulton's hopper; Fig. 9 are two 'forms 

 devised by Mr. Allen ; and Figs. 10 and 11 Mr. Battye's. Each 



