428 



SCIENCE 



[N. S. Vol. XLIX. No. 1270 



fowl, conducted for several years at the Rhode 

 Island Agricultural Experiment Station, there 

 has been evolved a new method for the de- 

 tection of those birds in a flock that are char- 

 acterized by higher producing ability. The 

 method is not based upon data involving ob- 

 served numerical production in any way, but 

 upon the tendency on the part of normal hens 

 to manifest, at certain periods of the year, 

 a gradual increase or a gradual decline in the 

 size and weight of the eggs which they lay. 



It has been found that when the numerical 

 production curve of a flock of hens of the same 

 approximate age and condition, and character- 

 ized by mediocre producing ability, is plotted 

 on monthly ordinates (aside from the mode 

 of December production of the pullet year 

 which is sometimes manifested if the hens 

 were hatched very early in the season or are 

 high producers) two modes appear, one in 

 April and one in September. These modes, 

 or maxima, represent the peaks of production 

 for the first laying year. 



It has also been found that when the curve 

 of mean egg-weight is plotted on similar 

 monthly ordinates, two modal points appear, 

 one in April and the other in September. 

 These weight modes, or weight maxima, are 

 approximately coincident with the production 

 maxima. 



"When, however, one analyzes the perform- 

 ance of individual birds at the period of these 

 maxima, one finds that, while the majority 

 show an egg-weight which has markedly in- 

 creased over the mean weight of the first ten 

 eggs laid at the beginning of the laying year, 

 some have not shown such an increase, and 

 some have shown, an actual decrease. When 

 the increase or decrease in mean egg-weight is 

 measured as a percentage-increase or as a 

 percentage-decrease, some birds may show an 

 increase of 10 per cent, or more while others 

 show a decrease of equal amount. 



The question naturally arose whether the 

 hens which showed the greater increase in 

 mean egg-weight for April or for September 

 also manifested the higher productions for 

 the first laying year. Such correlations were 



computed and it was learned that in the 

 majority of cases the hens which gave the 

 higher-percentage increase in egg-weight dur- 

 ing these months also showed the higher pro- 

 ductions for the year. The flock could easily 

 be separated into production-groups based 

 upon the percentage of increase (or decrease) 

 in mean egg weight. As a rule the correla- 

 tion was more perfect in September than in 

 April. 



Correlations between the percentage of in- 

 crease in mean egg-weight and niunerical 

 production were also attempted when the 

 former were based upon the mean weight of 

 only ten eggs laid as nearly as possible to the 

 absolute weight-mode for April and for Sep- 

 tember respectively. It was learned from 

 these computations that the correlation was 

 even more perfect when the smaller nmnber 

 of eggs was employed in the computations. 

 The following table gives the results for the 

 autumnal weight maximum, based on the 

 " 10-egg test." 



TABLE I 



Showing the Mean Annual Production, for the 

 First Laying Year, of Groups of Hens Selected 

 for Varying Percentages of Increase or De- 

 crease in Mean Egg-weight of Ten Eggs 

 or Less Laid at the Period of the Au- 

 tumnal Weight Maximum 



Percentage-Class : 

 Birds Selected 



for Increase in Number ol Mean Produc- 



Bgg-Welght In- Individuals tlon for the 



dicated Below. , Making the First Laying 



Per Cent. Record Year 



>13 3 147 



>11 6 145 



> 10 7 143 



> 9 8 144 



> 8 10 144 



> 7 12 139 



> 6 12 139 



> 5 14 135 



> 4 16 134 



> 3 19 131 



> 2 21 125 



> 1 23 125 



> 26 124 



.< 5 108 



< 6 19 112 



Total flock 31 120 



