268 HARRIS, KIRKPATRICK, BLAKESLEE, WARNER AND CARD 



of months and combinations of months in order (a) to determine the months 

 which give the best results and (b) to enable those who wish to predict 

 from any group of months. 



In the investigations one phase of which is presented in this paper, we 

 have sought among other things: 



(1) To determine the best method of predicting the annual egg pro 

 duction of a bird from the known record of any individual month. 



(2) To determine the best method of predicting the annual egg produc 

 tion of a bird from the combined records of two or more months. 



(3) To determine the best method of predicting the egg record of a hire 

 during a portion of the year from the record of a single antecedent montl 

 or a group of antecedent months. 



(4) To compare the relative merits of these several methods of predictioi 

 among themselves and to determine thereby which of the methods make 

 possible the most exact prediction as a basis for determining which is likel; 

 to be of the greatest practical value. 



We fully recognize, and desire to emphasize especially, the fact that th 

 whole problem of the prediction of future egg production cannot be solve 

 in a single investigation. The problem is exceedingly complex and 

 number of factors are not taken into consideration at all in the presen 

 paper. All that has been attempted is to indicate the possibility of a 

 important line of advance and to lay the foundations, in a series of statistic; 

 constants, for wider investigations. Some of these are already in progres 

 In the meantime, the results presented here may prove useful both froi 

 the practical standpoint and in facilitating to some extent further and moi 

 adequate investigations. 



The first definite step in the direction of the use of the egg record of 

 short recorded period for the prediction of the probable production durir 

 a subsequent or a longer period was, as far as we are aware, taken in 19! 

 when it was shown (HARRIS, BLAKESLEE, WARNER and KIRKPATRK 

 1917) that in a heterogeneous series of birds such as are submitted 1 

 practical breeders in egg-laying contests, the October egg production 

 correlated with that of every other month of the year. The whole subje 

 was carried much further in a second memoir (HARRIS, BLAKESLEE ai 

 KIRKPATRICK 1917, 1918) in which the correlations between the recor 

 of the individual months and the production of the whole year, betwe 

 the records of the individual months and of the remaining 11 months 

 the year, and between the production of 5 of the individual months ai 

 the production of all the other individual months, were published for to 

 series of birds. In this paper the equations for the prediction of toi 



