52 MAINE: AGRICULTURAL EXPE^RIMENT STATION. I903. 



about ventilation. If the closet was filled with water it would 

 quickly leak out. The cracks around the curtain frame, at top 

 and bottom, furnish openings for the entrance of fresh, and 

 escape of foul air. 



The main part of this house is cold of course, but the straw 

 on the f^oor is always dry, and the air clean. \Miile the birds 

 are on the roosts — in bed — they are warm. They come down 

 to their breakfast and spend the day in the open air. Such treat- 

 ment gives vigor and snap to the human, and it seems to work 

 equally well with the hen. 



EXPERIMENTS IN INCUBATION IN I902. 



In many experiments there are disturbing causes which inter- 

 fere with the work and render unreliable the results, which seem 

 to point in certain directions, and indicate certain truths. This 

 is true with all of the investigations we have made concerning 

 incubation. We have in use several good incubators and have 

 had considerable experience in operating them and others. We 

 think we have no trouble in getting as good results as when the 

 eggs are submitted to the careful treatment given by the mother 

 hen. With eggs from hens that have been laying but a short 

 time we usually get hatches of from 70 to 80 per cent from the 

 entire number of eggs incubated. 



During the last three years we have planned and carried to 

 completion a large number of incubation tests. By use of the 

 trap nests we were able to know the source of every tgg and it 

 was thought that the eggs from the same hens, all laid within 

 ten days of each other, and divided by selecting the first, third, 

 fifth, etc., for one lot and the second, fourth, sixth, etc., for the 

 other lot, would be nearer alike in freshness, fertility and germ 

 strength, than though they were selected from different hens, 

 however much alike such birds might be. 



In this way we could have all the eggs of each hen in a 

 class of her own — in two lots — and it was thought that the com- 

 parison of one hen's eggs with each other would yield more 

 reliable data than would the comparison of the eggs from differ- 

 ent hens. This assumption is true if the individual bird yields 

 uniformly fertile eggs throughout the test. As investigations 

 progressed we became aware of the great variation in the chick 



