on development of pattern in birds.



161



tied form of the second. The Red-collared and Forsten’s Lorikeets

nearly approach the first of these types, but have an additional

differently coloured belt across the front of the abdomen. The

colouring of the head is very frequently sharply defined and distinctly

separated from the remaining body pattern ; and, as I shall explain

later, I think this may have been brought about through the reduced

size of the feathers on that part of the body. A sharply defined

breast-belt is another extremely common feature, as also a distinctive

colour on the rump.


As distinct from the above types, which probably owe their

present more or less bright colouring to the agency of sexual selec¬

tion, there are very many groups of birds which are somewhat

conspicuously coloured on the under parts, but uniformly brown,

with dark shaft-streaks to the feathers on the upper surface. Now

it is evident that if even the Warblers were coloured above as they

are below, they would run the risk of speedy extermination by pre¬

dacious enemies, so that the colouring of their upper parts has

clearly been acquired, or, more likely, retained, for protective

purposes.


The simplest and most prevalent marking in birds is the

shaft-streak, and therefore I think it probable that it was the parent

of all other bird-markings. In its simple form it is fusiform (spindle-

shaped) or hastate (spear-headed), from the latter by an elongation

of the posterior angles it becomes sagittate (arrow-headed); then it

may curve round, forming a sub-marginal or marginal band or border

to the feather; and sometimes when a crescentic border is formed

the shaft-streak vanishes, as though the dark pigment had been used

up in the later development. Take, for instance, White’s Thrush as

an example. The fork of a sagittate shaft-streak, if continued to the

outer margin of the wing, would leave a triangular spot of the

original colour in the centre of the border, and this probably

developed the pale terminal spots commonly found on birds’ wings.


Now, by modifying the shaft-streak and the crescentic border,

I think almost any pattern to be found on birds may have been

developed. Of course, I am not speaking of secondary sexual char¬

acters affecting the form of the feathers, but only of the markings in

more normal plumage. There is no doubt that the narrow parallel



