670 PRACTICAL PROBLEMS 



This latter point is covered in the system in use by A. J. Love- 

 joy, of Roscoe, Illinois, a well-known breeder of Berkshires, a 

 sample card from whose herd is here reproduced : 



Index No. 16. Imported Bessie II 55101, farrowed April 10 to Master- 

 piece 77000. Farrowed 5, saved 5 : boars 4, sow i 



Sold boar to J. W. Martin, Gotham, Wisconsin . . . '. .'. '. .$150.00 



" " " L. W. Brown, Berlin, Illinois 75-oo 



" " " J. R. Logan, Seward, Illinois ......... 50.00 



" " " Hibbard & Brown, Michigan 125.00 



" sow " Nebraska Insane Hospital, Lincoln, Nebraska . . . 50.00 

 Total for litter of 1905 $450.00 



In connection with office records of this kind the "breeding 

 book " kept at the barn should record the date of every service, 

 the date, sex, and any distinguishing fact concerning the birth 

 of every individual, living or dead, and any other fact that would 

 prove of the slightest value in estimating what the herd is doing 

 or has done. 



If, then, in addition, there were kept an accurate " descriptive 

 record" of every individual that is considered worthy to enter 

 the herd, or to be sold as a breeder, and besides this also as ac- 

 curate a record, of the unworthy products of the herd, the breeder 

 would have not only for his own satisfaction and profit when 

 memory fails or becomes confused, but for that of others who 

 may handle his breeding a record of qualities good and bad 

 on which a skillful breeder can safely base his selection. With- 

 out a record such as this all real selection is limited to what is 

 done with the eyes on animals- still living, except as a man may 

 be guided by an uncertain memory. How uncertain that is he 

 will realize when he revisits in full prime of life the hills and 

 valleys of his boyhood. 



Pedigree records. That which is needed within the herd is 

 equally important within the breed. The facts of heredity go 

 to show that all good breeding requires that the type shall be 

 unchanged for at least six generations, if we hope to get any- 

 thing like uniformity of offspring. If this be true, we need 

 accurate records covering all important details and all valuable 

 characters for at least the six generations required to produce 

 a stable type. . 



