586 GENETICS IN RELATION TO AGRICULTURE 
3. The use as breeders of such males only as are known to be the sons of high- 
producing dams, since only from such males can we expect to get high-producing 
daughters. 
4. The use of a pedigree system, whereby it will be possible at least to tell what 
individual male bird was the sire of any particular female. This amounts, in ordinary 
parlance, to a pen pedigree system. Such a system is not difficult to operate. In- 
deed, many poultrymen, especially fanciers, now make use of pen pedigree records. 
It can be operated by the use of a toe-punch. All the chickens hatched from a par- 
ticular pen may be given a distinctive mark by punching the web between the toes in 
a definite way. If one desires to use a more complete individual pedigree system, he 
will find the system described in Bulletin 159 of the Maine Agricultural Experiment 
Station a very simple and efficient one. It has been in use at this Station for 7 
years, with entire satisfaction, on the score of both accuracy and simplicity. 
5. The making at first of as many different matings as possible. This means the 
use of as many different male birds as possible, which will further imply small matings 
with only comparatively few females to a single male. 
6. Continued, though not too narrow inbreeding (or line-breeding) of those lines 
in which the trapnest records show a preponderant number of daughters to be high 
producers. One should not discard all but the single best line, but should keep a 
half dozen at least of the lines which throw the highest proportions of high layers, 
breeding each line within itself. 
In the above set of directions two things will challenge the student’s 
interest most, namely the emphasis which is laid upon constitutional 
vigor and vitality in the selection of breeding birds, and the fact that a 
system of line-breeding or inbreeding is used in order to increase 
fecundity. 
The relation of the above directions to the genotypic behavior are 
not difficult to point out. Of females there are two different types 
(ZM)WLL and (ZM)WLI which are high producers; the remainder are 
either mediocre or low producers. It is assumed that by trap-nest 
records, it has been possible to segregate out a certain number of such 
high-producing hens from a mixed flock of low, high, and medium pro- 
ducers. When these are mated to males from the same lot, a variety of 
results will be produced according to the genetic constitution of the males. 
In Table LX XIX are collected the results which follow when females of 
the two high-producing genotypes are mated with the nine possible kinds 
of males. Now if the numbers of females of genotypes (ZM)WLL and 
(ZM)WLI in each pen are approximately equal—in practice those of geno- 
type (4M)WLI would probably be in excess—then it will be practically 
impossible to distinguish matings of types (1) to (3) and possibly (4) and 
(5) unless the number of daughters tested from each pen be relatively 
large. In this connection we recall the fact, as a further difficulty, that 
modifiability in egg production is relatively very great. If now an 
equal number of pens from matings (1) to (5) should happen to have 
Digitized by Microsoft® 
