GENETICS: HARRIS, BLAKESLEE AND KIRKPATRICK 567 
regression equations, which have been found to give reasonably good 
fits to the empirical means for the annual egg records of birds laying 
various numbers of eggs in the individual months. The slope of the 
lines when plotted shows there is an increase of from 2.6 to 5 eggs in 
mean annual production associated with a variation of one egg in 
monthly record. Since in practical selection groups of birds differing 
by far more than a single egg may be recognized, the difference in an- 
nual production secured by selecting in any month may be of very 
practical importance, amounting to from 30 to 60 eggs per year. 
FIG. 2 
While the correlation between the monthly egg record and the total 
annual production of the bird is the measure of the greatest practical 
value, forming as it does the basis for the prediction of average annual 
production from the observed performance of a limited period, neither 
it, nor the biologically more satisfactory correlation between monthly 
record and the production of the remaining eleven months, gives all 
