198 Earl W. Benjamin 



shell early in its development, while some is added later — as is indicated 

 by the lighter shade of an egg that has been laid prematm-ely, due to 

 some excitement. The intensifying of the pigment with the age of the 

 bird is supposed to continue until she has attained her full vigor, when 

 the tint begins to decline gradually. Newton believes that except for 

 individual differences the pigment is fairly constant in supply. 



Sorby (1875) found seven substances which in various mixtures are sup- 

 posed to produce all eggshell colors. These substances were oorhodeine, 

 oocyan, banded oocyan, yellow ooxanthine, rufous ooxanthine, a substance 

 giving narrow absorption-bands in the red, and lichnoxanthine. They are 

 said to be closely connected with either haemoglobin or bile pigments. 



M'Aldowie (1886) and many others have advanced theories as to the 

 cause of variation in eggshell color. The general opinion seems to be that 

 the color is very unstable and variations do occur frequently, and that 

 general tints or colors are inherited. Horwood (1909) gives it as his 

 opinion that coloration of the shells of birds' eggs has absolutely no 

 connection with mendehan principles. 



According to Surface (1912), the color of eggshells is probably added 

 from glands in the vagina or adjoining parts of the oviduct, and it may 

 reasonably be supposed that a fimction of this nature would be inherited. 

 Such a supposition agrees with the results of Benjamin (1912 and 1914), 

 which are discussed later in this report. 



All these studies, made by various workers, show conclusively that 

 with respect to ma,ny characters, including size, shape, and color, there 

 is a characteristic type of egg to be accredited to each individual, and 

 that some degree of inheritance has been found to exist. 



n METHODS OF INVESTIGATION 



The investigation described in this memoir was begun, in the spring 

 of 1911, by selecting fifty eggs for hatching for each of the follow- 

 ing nine characters — three characters being grouped in each of three 

 selection studies: 



