132 STUDIES IN THE EVOLUTION OF ANIMALS 



again turn up as atavic marks, perhaps by some atomic disturbance of 

 crossing in the nerve-centres. 



The curious point is, that white, black, and tan colours are interchangeable 

 — why, I do not know. This can be readily seen in the Fox-Terrier. The 

 black patches of the one are almost identical with the tan patches of the 

 next. The black and tan colours of the one become tan and white 

 colours of the next, and so on. Even the tan spots over the eyes have 

 been known to change into white spots. So that it is no wonder that we 

 see black spots or stripes in one order of mammals changing into modified 

 white spots, or modified white stripes, in another order of mammals. 



The interchangeableness of these three colours, with variations in the 

 shades of the tan, can be seen in many other mammals, both domesticated 

 and wild. 



Lastly, when all vestiges of rosetting, spotting, or striping have dis- 

 appeared, as in self-coloured mammals, there may remain a vestige of the 

 ancestral carapace as a whole, without any vestige of the separate plates. 

 This I take to be the meaning of the dark colour of the dorsal and flank 

 regions, as contrasted with the abdominal coloration, which I take to be a 

 vestige of the ancestral unarmoured surface. The contrasted colours may 

 be black, tan, brown or white (brown being a combination of tan and 

 black). 



That curious fiank band which some mammals have, clearly seen in 

 certain Squirrels, may possibly be a vestige of the receding carapace in 

 process of total obliteration, and the different contrasted colours may in- 

 dicate the different ages of partially unarmoured surfaces — that is, the 

 abdominal colour would indicate the oldest unarmoured surface, the 

 flank band a later unarmoured surface, and the dorsal colour the latest of 

 all; The origin of this flank band is, however, not so clear to my mind 

 as. the other parts of this theory. I think we will find it impossible to 

 account for every line, or band, or spot in the coloration of mammals, 

 because the modifications in their coloration have been very great. 



We cannot call the whole self-coloration as the youngest of all, for 

 there may have been selfs among marsupials very much earlier than 

 arnor;g other orders. Each group of mammals should, I think, be studied 

 by itself, with an eye to its descent from some sort of carapaced ancestor. 



It is a marvel to me that so much has been left to us of these hiero- 

 glyphics in \yhich ancestral characters had been written, and that they 



