572 EXPERIMENT STATION RECORD. 



Station in the experimental breeding of poultry with reference to egg produc- 

 tion, with some additional data not hitherto published regarding obscure and 

 doubtful points of interpretation. 



It is concluded that trap nesting for the purpose of improving egg produc- 

 tion by the selection of the best layers has not that degree of practical useful 

 ness and importance which it was popularly supposed to have some 10 years 

 ago. Its only profitable function in practical or commercial poultry keeping, 

 as distinguished from experimental, is deemed to be in connection with special 

 needs or problems, as in the work of the fancier who desires to keep individual 

 pedigrees of his stock. 



In the interpretation of earlier work the author finds that the percentage of 

 mortality fluctuated in amount from year to year. These variations are ac- 

 counted for by differences in general environmental factors and to accidents. 

 There was a general trend downward, however, during the period 1S99 to 

 1908, when egg pi'oduction was showing a slightly downward trend. There is 

 no evidence that either artificial incubation or any other environmental con- 

 dition affected adversely the constitutional vigor of the strain. The percentage 

 of infertile eggs was distinctly high, because the eggs for hatching were used 

 soon after mating, being 19.3 per cent in females of high fecundity lines and 

 18.8 per cent in those of low fecundity. The hatching quality as indicated by 

 the percentage of fertile eggs hatched was somewhat below what may be con- 

 sidered normal for the station's stock at the pi-esent time. Taking all the 

 records together, in the high fecundity lines it required 2.6 eggs in the incubator 

 in 1911 to produce 1 chicken 3 weeks old, while in the low fecundity lines it 

 required 3.2 eggs to make 1 chick 3 weeks old. There was no substantial dif- 

 ference between the females of the high and low fecundity lines in respect to 

 hatching records. 



An experiment in out-crossing, involving a large number of individuals, indi- 

 cated that the infusion of new blood failed to produce any change in the egg 

 production of the progeny. Therefore, it is concluded that the amount of in- 

 breeding practiced during the mass selection had no unfavorable influence on 

 either egg production or on the general vitality of the stock. As there was no 

 increase in egg production by mass selection in the laying year 1907-8, a new 

 plan of breeding was adopted as a working hypothesis, having for its basis the 

 genotype concept of Johannsen, which is explained in detail. 



Data are presented showing that it has been possible to isolate and breed 

 stock strains or lines in which high fecundity has not been changed during 4 

 generations at least. This indicates that high fecundity and low fecundity 

 segregate in accordance with Mendelian principles. 



Poultry notes, 1910, R. Pearl (Maine 8ta. Bui. 193, pp. 111-200, figs. 8).— 

 A report of progress in the poultry work at the station in 1910. 



The brooder which had been used proving unsatisfactory, a new type was 

 devised which proved satisfactory. A description and working plans of the 

 new brooder are given. 



Experience has shown that the roosting closet was of no advantage and has 

 been abandoned. 



A study of the residual error of trap-nest records is presented in detail. 

 With the old trap nests the residual error was 4.01, with the new 1.24 per cent. 

 The relative amount of unrecorded egg production was not closely related to 

 the total egg production. The absolute number of unrecorded eggs tended to 

 increase as the nest eggs increased. During the first 2 years the tendency for 

 the unrecorded production diminished relatively. The longer the same in- 

 dividual birds used the trap nest the smaller became the production of im- 

 recorded eggs, which suggests that there is an element of learning in the 



