Essay on the Interpretation of Milk Records. 169 



The most important calculation will be to estimate the 

 standard R. M. from first calf R. M. The lai'ge variation 

 coefficient of the latter shows this to be considerably leps 

 accurate than estimation from second calf R. M., but the great 

 saving in time effected by judging the cow after the first calf 

 would probably financially compensate for this loss of accuracy\ 

 More extensive data are being employ^ed for the calculation of 

 correlation and regression coefficients between R.M.,, R.M.j, 

 R.M.3, and R.M.g, and the present number of cow^s does not 

 justify the use of an exact factor ; but it will be seen that the 

 following rough corrections bring the means within a quart of 

 the standard R.M. : — 



Mean R.M.j + 50 per cent. = 14-7. 



„ R.M. 2 + 20 per cent. = 15-4. 



„ R.M., + 10 per cent. = 1.5 8. 



„ R.M.9 = 156. 



Now if suitable corrections are made for exterior circum- 

 stances affecting the R.M., the variation found in the latter 

 from year to year with the same cow should be diminished. 

 Frequency curves have therefore been plotted of the " range " 

 of R.M. for the 110 cows for six years. For example, an 

 individual cow giving successive R.M., 8 — 10 — 10 — 12 — 14 — 

 13 is entered as having a range of 6 quarts. When above 

 corrections for age are made, her R.M. become 12 — 12 — 11 — 12 

 — 14 — 13, and her range is 3 quarts. 



The frequencies of the range of variation found in each of 

 these 110 individual cows during the fii'st six years of their 

 milking career yielded the following constants : — 



After correction for age by the addition of 50 per cent, to 

 first calf R.M.i, 20 per cent, to R.M.j, and 10 per cent, to R.M.3, 

 these were reduced as follows : — 



Mean of individual " ranges " corrected for age 502 quarts. 

 Median „ „ ,, „ 4-67 „ 



Mode „ „ „ „ 3-97 „ 



Influence of season of year of calving. — For this point the 

 extracted records of two farms were taken, which amounted to 

 1,418 lactations. 



The mean R.M. for cows calving in each month of the year 

 was calculated. These are given, together with the number of 

 records from which they^ were obtained, in column 1 of the 

 following table. In column 2 is given the deviation of each 

 monthly mean from the mean of the whole ; and in column 3 



