166 CALIFORNIA ACADEMY OF SCIENCES. 



Olid feather has two complete bars and a triangular sub- 

 terminal dot. The tendency, however is towards a more 

 complete pigmentation, and in the third feather the 

 bars have run together somewhat along the line of the 

 shaft. The triangular patch has become extended up- 

 ward along the shaft also. In the fourth feather this 

 darkening of the feather has proceeded another stage, 

 while in the last of the set (fig. 5), the outer web is 

 almost entirely black, and the inner web has succeeded 

 in closing in an island of light, a tendency which had 

 been developing through the whole series. In the 

 primary coverts no such transition is observable except 

 the gradual loss of one of the bands (figs. 6-15). The 

 primaries themselves, however, present an interesting 

 series of changes (figs. 16-25). 



In the first place it will be noticed that the number 

 of bands is largely dependent upon the size of the 

 feather. If there is a little more room at the basal por- 

 tion another bar Avill be formed. In the first feather 

 (fig. 16) there are eight white bars. In the next two 

 which are the longest there are nine (counting the spots). 

 In the fourth (fig. 19) there are only eight, this number 

 persisting in the fifth and sixth, although the basal bar 

 is very indistinct. In the seventh feather (fig. 22) there 

 are but seven bars, this number being reduced to six in 

 the last three — and the sixth one very inconspicuous. 



It seems to me we have here a fact of considerable 

 significance. This basal bar, or even two or three 

 of them perha])s, could never be of any use because 

 they would be concealed in life by the coverts, and must 

 therefore be repeated from some general law. Just as 

 soon as there is a little space at the base of the 

 feather another bar of pigment crowds in. It seems 

 probable that this is to- be accounted for by the law of 

 repetition, so closely akin to the law of correlation. In- 



