156 Maine Agricultural Experiment Station. 1920. 



the butter-fat percentage to be expected for the first five lacta- 

 tions milk of cows which have the butter-fat percentage of those 

 shown in the first column. 



Noting the expected butter-fat percentage of the highest 

 and lowest tested cows in this column it is seen that those cows 

 which test very low in butter-fat (3.75 to 3.85) in their first 

 lactation on the average increase 0.581 per cent of butter-fat for 

 the milk of their first five lactations ; for the highest butter-fat 

 test cow in their two year lactations the butter-fat percentage for 

 their first five lactations is on the average 0.645 P er cent °^ but- 

 ter-fat lower than is the test for the first lactation test. 



The third and fourth columns give the summations of the 

 number of cows from the lowest butter-fat test to the highest 

 and from the highest butter-fat test to the lowest. The last two 

 gives the butter-fat test expected for these one thousand two 

 year old Jersey heifers when the herd is divided at any given 

 place. To illustrate suppose the owner is to cull this herd so 

 that only cows which produced 5.25 per cent of butter-fat in 

 their first lactation will remain in it, what per cent of butter-fat 

 could he expect the remainder of the herd to produce? The 

 answer is found in the last column of the table on the line with 

 the 5.25-5.35 butter-fat percentage of the first column, to be 

 5.49 per cent. Should it have been desirable to know the butter- 

 fat percentage of the culled section of the herd this is found in 

 the sixth column to be 4.97 on the line with 5.15-5.25 of the first 

 column. The number of individuals remaining in this herd after 

 culling may be seen in the fourth and fifth columns. Since these 

 are tabled for 1000 individuals they may be easily reduced to 

 percentage should it be convenient to deal with the results in 

 this way. 



These results are of course only applicable to herds similar 

 to the one being studied. The comparison of the butter-fat test 

 make it seem likely that with only small error these results may 

 be used for the whole of the Jersey breed and possibly the 

 Guernsey breed. 



