EVOLUTION OF COLORATION. 227 



laid clown in ontogeny, and hence mature first. This phenomenon of priority 

 in development is not at all of a phylogenetic nature; it allows, I believe, of 

 no other interpretation than that it is simply a phase of the general law that 

 the anterior and proximal portions of an animal develop ahead of and faster 

 than the posterior and distal. In variation the posterior distal and dorsal are 

 most variable, as they are the last developed, and are hence the longest sub- 

 jected to the causes of somatic variation when in an undifferentiated state. 



(2) Color patterns. The color patterns of the body are built up of color 

 areas derived from the phyletic system, of segmental centers of coloration, 

 which appear in the embryo. Of these centers the species uses those which 

 its peculiar constitution demands and suppresses all others. Upon the wings 

 color appears in centers, but each species in the ontogeny of the wing develops 

 pigment cells in the centers where the constitution of the species demands 

 them. In general these are developed in stripes, bands, or spots, located at 

 very definite and invariable positions upon the wing. Stripes are always 

 found between the veins and bands at right angles to them (plate 24, fig. 2), 

 and at the following places: Base, apex (margin), middle, above the trans- 

 verse vein, one half-way between the basal and middle bands, and another 

 half-way between the middle and marginal. Spots occur most frequently at 

 the point of intersection of bands and stripes. 



(3) Color characters. In development and evolution the units of color- 

 ation upon the body are the segmentally placed centers, while upon the wings 

 they vary in different races, but are in general whole bands or stripes, or por- 

 tions thereof. 



(4) Color-pattern evolution and color variations, both somatic and germi- 

 nal, coincide, and are orthogenetic upon the body because of the existence of 

 the system of phyletic color centers, which follow in variation and evolution 

 the laws governing metamerically repeated structures, and on the wings 

 because of the phyletic system of stripes and bands which follow their own 

 laws. 



(5) Color-pattern evolution tendencies exist which produce patterns of 

 definite type due to some force characteristic of the germ plasm. This force 

 takes certain color centers from the phyletic system in the embryo, and by 

 various modifications thereof produces diversity in coloration. In Leptino- 

 tarsa there exists a general generic tendency which controls the entire genus, 

 a group tendency which governs groups of species, and a race tendency which 

 is limited to a few species. These tendencies are not to be confounded with 

 the "internal perfecting principle" of Nageli and others, but they are entirely 

 a matter of germ-plasm constitution and of heredity, and any change of the 

 latter will modify the tendency. The difference betwen these two ideas is 

 this : the "internal perfecting principle" involves the conception of an ideal 

 for the race toward which it is driven by the directive force of some more 

 remote agency ; while a "tendency," as I use the term, is an expression for a 



