>I have been reading a book by Michael Denton which talks 
>about some of the problems with wing development in the evolutionary 
>scheme of things.
>
>Quoting from Evolution A Theory in Crisis page 202:
>
>Take, for example, the problem of the origin of birds. The flight feather 
>of a bird is one of the most beautiful and well known of all biological 
>adaptations.
>
>Each feather consist of a central shaft carrying a series of barbs which 
>are positioned at right angles to the shaft to form the vane. The barbs 
>which make up the vane are held together by rows of barbules. From the 
>anterior barbules,hooks project downward and these interlock with ridges 
>on the posterior barbules. Altogether, in the flight feather of a large 
>bird, about a million barbules cooperate to bind the barbs into an 
>impervious vane... 
>
>There is much more describing the so called Handley Page slot, design of 
>aerofoil, provisions of slots all of which the birds wings exploit. The 
>intricate system of tendons which allow the feathers to twist in such a 
>way that when the wing is raised they open like the vanes of a blind, 
>greatly reducing resistence, but close completely on the downstroke, thus 
>greatly improving the efficiency of flight.
>
>If as I understand birds were supposed to evolve from reptiles, and that 
>the feather evolved from a reptile's scale, I find it difficult to see 
>what other group of living organisms could possilby serve as hypothetical 
>ancestors. 
>
>Again Menton says: By what sequence of events and through what kinds of 
>transitional states might the feather, the feathered aerofoil or wing, 
>and avian flight have evolved?
>
>I think it's a good question.

Just thought I would look up the development of feathers.  I have an old
developmental biology book (but since this is an old question, I don't
have any problem with that): _Chordate Structure and Function_ by Allyn
J. Waterman,  McMillan Co. 1971. p. 156-62.  To quote (not out of context):
"The early developmental stages of a feather and a scale are similar - 
both show a vascularized dermal core covered by a thickened layer of 
ectoderm.  As the feather papilla lengthens into a cylindrical cone, it
grows downward into the dermis and forms a pit around its base which
continues to deepen into the follicle.  Concomitantly the ectodermal
component of the papilla grows outward, to form the keratinized quill
enclosing the dermal pulp, outer sheath, and subsequently the bundle of 
ectodermal barbs. As the sheath splits and the barbs become separated 
and spread flat, the vane of the feather is produced."
In short, feathers ARE highly modified scales.  Looking at a more
recent developmental biology book, _Developmental Biology_ by Scott
Gilbert, Sinauer, p. 560, my edition is the Second (1988) but I am sure 
there is a more recent one, you will find that feathers can be induced by
mammalian dermis, which normally induces hair (also highly modified
scales).  This is not some wild idea by evolutionary biologists.
Developmentally, there are many similarities among hair, scales, and
feathers.  There would be even in "Creationworld".  But they
would be unable to answer the question of why this should be the
case.
To further quote Waterman:
"Developmental patterns in all kinds of feathers are differentiations
from a common basic type."
In short, don't look *just* at the most extreme modification of a feather
and claim it is impossible to reach that extreme.  It is not necessarily
impossible to reach an extreme by modification of other, simpler feathers.
If you ignore all the intermediate steps present in other feathers
it might be impossible to explain as one swell foop.  But it gets pretty 
easy if you include possible precursors.  
In fact, these simpler feathers often serve insulating functions.  This
is what leads many to think feathers were initially developed not for
flight, but for warmth.