no MAINE AGRICULTURAL EXPERIMENT STATION. I909. 



of the other. Nowhere in this paper are the statistics of these 

 two years lumped together. So long as the figures for the two 

 years are not lumped together and conclusions are not drawn 

 which depend upon joining ^ one of the years with the 

 other, it is a great advantage to have the conditions so different 

 in the two years. It affords an opportunity to determine 

 whether relations which obtain under one set of conditions will 

 hold under a totally different set. 



The hatching in all of this work was, as in previous years, 

 done in incubators. All incubators used were Cyphers No. 3 

 (360 ^g'g) machines. The only methods used which differed 

 at all from those previously followed in the Station's poultry 

 work were such as were necessitated by the fact that pedigree 

 records were kept."^ The eggs from the different birds were 

 all handled in the same way and subjected to the same condi- 

 tions so that differences in hatching results can not be referred 

 to the treatment which the eggs received. It is conceivable 

 that the hatching records as a whole might be improved or 

 made poorer by using some other methods of handling the eggs 

 during incubation. But it is not conceivable that when all eggs 

 are handled in the same way differences between the eggs of 

 two individual hens in regard to hatching qualities can be 

 explained as the result of the handling. Such differences are 

 inherent in the eggs themselves. It is with differences of this 

 kind that this paper has to do. Care was taken to insure that 

 the eggs from given individual hens should be distributed at 

 random through the different incubators, in order to guard 

 against any possible inequality of treatment from this source. 



The eggs were tested for fertility by candling in the usual 

 way. This testing was done by Mr. Walter Anderson. His 

 long experience in work of this sort insures the substantial 

 accuracy of the fundamental data. In the 1908 w'ork a check 

 was kept on the determinations by opening eggs at intervals. 

 It is not contended that the records of infertile eggs of that 

 year include absolutely none that had begun development and 

 stopped very early. Such absolute accuracy is unattainable 

 wdthout opening every egg. The number of errors of this kind 

 in the records we know to be very small, however. There can 

 be no question that they do not in any way affect the results. 



Cf. Me. .\gr. Expt. Stat. T'.ullctin 150. 



