78 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. xxx. 



the sides gives way to separate spots (Fig. 2, Plate X), but this is not 

 an essential change as some specimens show a nearly continuous bar 

 from back of the eye to the middle of the tail. The ground color, 

 however, graduall}- deepens, becoming vermilion in highl}- colored 

 specimens. 



The development of the color pattern mu}^ be summed up as follows: 

 Originally pigment cells are evenl}'^ distributed except for a ventral 

 and three lateral series of pigmentless areas. In later development 

 these acquire pigment by influx from adjoining parts thus causing a 

 dilution of pigment. Otherwise the keynote of the entire process 

 is concentration. 



The pigment of the head, legs, and tail moves little, simply crowd- 

 ing together in blotches. The concentration of the pigment cells of 

 the tail presents one notable feature, namely, that a few cells remain- 

 ing in the keel are laid down as median spots when that portion is 

 absorbed, forming the onl}' noticeable group of so located spots on 

 an}^ part of the animal. 



The movements of pigment cells on the body proper are more com- 

 plex. The spots on the dorsal surface are formed of pigment from 

 the immediate vicinity and also from the border of the upper series of 

 immaculate areas. The fact that the first spots originating here are 

 lateral pro])ably accounts for the scarcity of true median spots. The 

 large amount of pigment paralleling the second series of pigmentless 

 areas on each side collects into the two series of distinct blotches on 

 the side of the adult. The pigment of the third series forms many 

 small scattered spots on the flanks of the adult and contributes to the 

 pigmentation of the ventral area. 



While in the development of the pigment pattern there seems to be 

 no increase in the number of cells, the change of the ground color 

 from l)ufl' to orange and vermilion is accomplished wholly 1)V the 

 addition of coloring matter. 



MELANI8Tir SPECIMENS. 



A point that arises in connection with the matter of the addition of 

 pigment is whether the so-called melanistic specimens really have an 

 extra amount of pigment, or whether their peculiar coloration mav 

 not be explained in another way. Dr. C H. Eigenmann gives an 

 account of two specimens of the cave salamander which have more 

 than the usual proportion of the surface pigmented. In one from 

 Rock House Cave, Missouri, there is a lateral streak ''broad enough 

 to cover the sides with a mottled pattern." Another specimen among 

 twelve from Marble Cave, Missouri, is described at length:'' 



The lower surface of the head i8 more densely piguientetl than in the otlier speci- 

 mens. The sides are more uniformly pigmented than in the melanistic individual 

 from Rock House. The sides of the head, body, the arms, and anterior surface of 



«Biological Bulletin, IV, 1903, pp. 227-8, fig. 1. 



