120 MAINE AGRICULTURAL EXPERIMENT STATION. I909. 



2. The mean or average percentage of fertile eggs hatched 

 in the 1908 season is seen to be approximately 37. This again 

 is an unduly low value. One has a right to expect a consider- 

 ably better hatching quality of eggs than this. Like the poor 

 average record for fertility it is, however, to be explained 

 chiefly by the unsuitable housing conditions under which the 

 birds were kept in that year, and in small part by the fact that 

 only pullet records are available. This is shown to be the case 

 by the 1909 figures where we have the percentage of fertile 

 eggs hatched increased to 51 per cent in round numbers tak- 

 ing all the birds of the season together. This last figure would 

 again be increased if the records of the early part of the breed- 

 ing season were omitted. It is believed, furthermore, that by 

 selection of breeding stock on the basis of hatching records it 

 may be possible to improve the average hatching quality of eggs 

 still more, (cf. discussion of this point in summary). 



3. The degree of absolute variability as measured by the 

 standard deviation is seen to be in all but one case ( 1909 

 yearling hens only) somewhat greater in the case of hatching 

 quality than in the case of fertility. The standard deviation is 

 a precise measure of the degree to which a group of individuals 

 conform to a type with respect to any character under investi- 

 gation. The closer to type a given lot of individuals run the 

 smaller will be the standard deviation exhibited by that lot. 

 On the other hand the more widely the individuals are scattered 

 about the type the larger will be the standard deviation. It 

 is not necessary in this place to go into the matter of how the 

 standard deviation is calculated. It suffices to say that it is a 

 scientifically accurate measure of the degree of closeness to 

 type. 



4. A\'hile hatching cjuality appears from the present statis- 

 tics to be absolutely a slightly more variable character than 

 fertility, if we consider the degree of variation in proportion to 

 the mean, the opposite is the' case. This is shown if from 

 Table IV the percentage which the standard deviation is in the 

 mean is calculated for each of the two characteristics. Per- 

 forming this operation the following results are obtained : 



1908 — Fertility: percentage of standard deviation in mean 

 = 69.1%. 



