252 THE AMERICAN MIDLAND NATURALIST 



present if not the original (juvenile) ground color of the entire shell. 

 In most cases this color was confined to the anterior and inferior 

 portions of the shell. The secondary color was that most evi- 

 dent on the superior and posterior portions. If present at all, it 

 was usually, but not always darker than the primary color from 

 which it was derived, and general observation showed that the 

 regions of the shell where it was found, to be the place of transition 

 from the primary color to it, whether to lighter or darker shades. 

 Where there seemed to be doubtful relationships between these two 

 recorded colors, two or three additional ones were taken for the 

 purpose of tracing genetic relationship in the sequence of color 

 changes. The colors then judged most to match the Ridgeway 

 vStandard Colors were then written in figuring books opposite cal- 

 culatings made for the morphological features of each shell. Only 

 one color w^as usually taken in consideration when effort was made 

 to trace the sequence of change of color, but in view of the fact 

 that color variation might be traceable as stated above to various 

 influences, it was thought well to possess data which would serve 

 to balance the conclusions. In PL III, (whose synthesis is afterwards 

 described), two colors are given, the first is the secondary color; 

 where only one is given, it represents the sole color determinable. 

 Further, as the specific descriptions given of some of the shells 

 indicate that the umbo may be lighter colored than the rest of the 

 shell, some confusion may arise when in carrying the writers scheme 

 in mind, it is observed that in Pi. Ill some of the secondary colors 

 are lighter than the primary ones, as obtained by the above pro- 

 cedure. The general plan when the tables of distribution of color 

 were prepared, was to give preference to that color most impressing 

 the eye with its preponderance or vividness in the epidermis. 



The prevailing color then being alone taken into consideration 

 for the calculations, tables were prepared in the following way. 

 Where a large number of shells from one locality were concerned, 

 it was the usual practice to group all shells of a closely similar epi- 

 dermal color and compare as a whole with the shades given in 

 Ridgeway. By this means a general or average hue was obtained, 

 not accurate of course for every shell, but very convenient in 

 determining the "relative colors" of the shells at that locality 

 to ones near it, above or below in the particular body of water. 

 As a rule these relative colors were taken from a fairly large number 

 of shells, although a lack of material often compelled the use of 



