92 Scientific Proceedings^ Royal Dublin Society. 



advance made in America in breeding liigii-yielding dairy stock, and 

 advantage was taken of the meeting of the British Association at Winnipeg, 

 in 1909, to make inquiry into the question in Canada and the United States. 

 Herds were visited at a nnmber of colleges and exj^eriment stations as well 

 as at some farms in both countries. But here again there was little more 

 information to be gained. The records were not old enough. Besides, 

 many of tlie American records are confusing, for the reason that they usually 

 state the butter-yield rather than the milk-yield for a lactation or for a year. 

 The butter-yield depends mainly upon two factors, which are inherited 

 independently, namely, the milk-yield and the proportion of butter-fat it 

 contains. Consequently a statement of butter-yield conveys no accurate 

 information either as to the yield of milk or as to its qualitj' ; and, as a guide 

 in breeding, this method is bound to be elusive and uncertain, since the breeder 

 is not clear as to whether his cows are yielding milk in quantity or in qualit3'. 

 If the branches of a bank report to headquarters the value each has in 

 hand in precious metal, the officials at headquarters will easily add up and find 

 the total sum, but they will have difficulty in determining how much is gold 

 and how mucli is silver. 



It was also found both in the States and in Canada, but especially at 

 colleges and experimental-station farms, that stock-breeders are much 

 influenced b_y the wedge-shaped shoulder idea: the cause lying to some 

 extent, perhaps, in their sj'stem of judging stock by scoring card. So 

 strong is this influence that sons of a narrow-shouldered, low-yielding cow are 

 frequently ))referred as stock-bulls to sons of a higher-yielding cow whose 

 shoulder is broad. 



The American visit having brought the problem no nearer solution, it 

 was next arranged through the Department of Agriculture and Technical 

 Instruction to have a set of forms on which were printed questions as to the 

 yields in herds of cows, and as to the cows' parentage. These forms were 

 sent to breeders in the United Kingdom who were known to have kept 

 records, and who might be in a position to give information. A number 

 were kind enough to fill up the forms, and among them were found some 

 whose herds were large enough and wliose records were old enough to 

 supply data suitable for the purpose in hand. The answers showed, however, 

 that a considerable amount of preliminary investigation was necessarj' before 

 the data could be made use of. Breeders had been asked to give the records of 

 each cow for as many years as possible and to state the cows' ages against 

 their records for each year. They were also asked to state over against each 

 record the length of the lactation concerned. From the answers to these 

 questions it was seen that in order to compare each cow with the others, it 



