1910.] Some Aspects of the Science of Breeding. 715 



operations among the herds in question, we suggest that a 

 large amount of "experimental error" ought to be allowed 

 for; experience, no doubt, will lead to better methods and the 

 creation of more perfect instruments. Even then, however, it 

 would seem difficult, without a post-mortem examination, to 

 ascertain how much of the difference obtained by measurement 

 is due to the frame and how much to the condition of the 

 animal under observation. 



There are other considerations which we find of value when 

 dealing with the measurements taken to enable the above 

 diagrams to be made. As has been said, all the cows 

 measured are in Lord Rayleigh's herds. They have been bred, 

 or in a few cases bought, fed, and managed, under the most 

 skilful supervision, with the one idea of milk-production. It 

 is quite possible that many of the animals which are in Class B 

 are there from some accidental cause, and that such cows, had 

 circumstances been different, might have been promoted into 

 Class A. We must also emphasise the fact that the cows in 

 Class B, though the very worst in these herds, can only be 

 called bad milkers when contrasted with those in Class A. 

 Were it possible to get the records of some of our beef-breeds 

 it would probably be found that, by comparison, the cows in 

 Class B would be quite good milch cattle. 



There is another matter which seems worthy of considera- 

 tion, and that is, that though a cow may milk well when all or 

 many of the milk-indicating points are absent, yet it is possible 

 that she may be a less useful animal because of their absence. 

 This may be illustrated by only one example — for this subject 

 opens up a very vast field for speculation — a flat-sided cow 

 which milks well may, owing to the restriction of space in the 

 body cavity, be unable to carry a good calf, and consequently 

 fail as a breeder. 



It is not in any way intended to suggest that the judge in 

 the show ring or the buyer of pedigree stock at the auction sale 

 will be benefited by obtaining such measurements. It is 

 rather, on the contrary, suggested that a very thorough and 

 complete investigation should be made of the methods em- 

 ployed by judges at present, so that their practice may be more 

 ; completely understood. It is, however, the object of this 

 ! article to suggest how much the agriculturist may have to do 



