106 



The Bulletin, 



it down in his notebook. By doing this 12 days a year, lie is able to furnish 

 me with a record of the amount of feed which each cow in the herd has con- 

 sumed. He also weighs the milk from each cow night and morning, tests it 

 for butter-fat, and furnishes me at the end of the year vrith a record of the 

 milk and butter-fat each cow has produced. With this record as a basis, a 

 man can make intelligent selection of the cows for breeding purposes, for it 

 tells how much milk each cow has produced, how much butter-fat, and the 

 cost of producing a hundred pounds of milk or a pound of fat. or, in other 

 words, how much each cow charges for producing milk and butter-fat, and 

 how much profit she pays in a year. 



When we first started this cow-testing association work we wanted to know 

 how accurate it would be to weigh the milk and test it once a mouth ; so I 

 went to Minnesota, where at the Experiment Station farm they had been doing 

 this work for years. They gave me access to their records, and I selected 

 one day each month, and computed the average for a year. The following 

 table shows how close we came to the actual amount of production by follow- 

 ing this system of a test once a month; and you will all agree with me that 

 this is close enough for all practical purposes. The table shows the maximum 

 I)er cent of variation for any one year and the per cent of total difference 

 from the actual yield for nine years. 



VAIlIATI0^'S OF ESTIMATED YIELD EKOM ACTUAL YIi:LD OF MILK AND BUTTEK-FAT. 



Cow 1: 



Maximum variation. 

 Total difference 



Cow 2: 



Maximum variation- 

 Total difference 



Cow 3: 



Maximum variation, 

 Total difference 



Cow 4: 



Maximuna variation. 

 Total difference 



Cow 5: 



Maximum variation. 

 Total difference 



Cow ti: 



Maximum variation- 

 Total difference 



Cow 7: 



Maximum variation- 

 Total difference 



In the method used by the cow-testing associations the total difference for 

 a nine-year period is in no case over 2 per cent from the actual yield. 



When we consider that the cow-testing association method means the weigh- 

 ing and testing of the milk .lust one day a month, and that the results are as 

 close to the actual as above stated, we must conclude that the records of 

 performance as found by the cow-testing association method are sufficiently 

 accurate to enable the dairyman to weed out his unprofitable cows. 



In every herd there are some cows which do not make any profit. The 

 thing to do is to find out which are the ones. I liave had farmers say to me, 

 "Do you suppose I milk my cows twice a day. 3<>.") days in the year, and do 

 not know the good ones, and whether or not they pay for their feed?" But 

 I merely say I do not believe he does, even though the farmer milks the cows 

 every day. One cow comes in fresh and gives 40 or 50 pounds of milk a day, 

 and the farmer wonders at and remembers the great milk flow she gave when 



