THE BIRDS OF AUSTRALIA. 
The next step was the proposal of a distinct red form as a new species, 
Mirafra woodwardi, from Onslow, West Austraha. This very striking aberra¬ 
tion was figured also as a tliird species of Mirafra ui Australia. 
The correct valuation of these forms was recognised by Hall, who discussed 
eight specimens from Broome, N.W.A., and wrote: “ There is no rufous 
colouring, with only the palest isabeUuie over portions of them. Consequently 
these birds appeal to me as representing a subspecies. It is a well-loiown 
fact that certain desert birds distmctly guard themselves m the matter of 
coloration by agreement with their surroundings. My correspondent, iir. 
J. P. Rogers, has collected these particular eight birds on groimd that has 
the birds assimilating in colour with it. He further states to me that on a part 
of the Fitzroy River, N. W. A., where the gromid is greyish, the birds are greyish, 
but that the majority on the Fitzroy are chocolate, because the ground is of 
that colour. On the Robinson and Meda Rivers, Mr. Rogers further states, 
the birds are brown in agreement with the broAvn soil. On the Ashburton 
River, near Onslow, from w^here Mr. A. W. hlilligan received his type {M, ivood- 
wardi), the gromid is very rufous, and so red and friable that Mr. Rogers has 
seen a cloud of it above the river when he was fifty miles away from it. As 
the colour of the ground so is the colour of the Mirafra. Accordingly, a 
knowledge of the large areas of varyuig soils and the Mirafra associating 
upon them should, in my opinion, give us a complete and proper list of this 
genus, with all hut the type ranking subspecifically. (The italics are mine.) 
The light-coloured specimens which I shall refer to as M. horsfiddi 
pallidus. . . .” The name selected had been previously used so I renamed 
it M. milligani, but about the same time, but published a little earUer, 
Bianchi had altered it to M. horsfiddi halli. 
When Ingram examined the collection made at Alexandra he noted these 
were rufescent, not pallid, so named tliis form binomiaUy M. mfescens. 
HaU was quite correct in anticpating the reduction of ah the forms to 
subspecific rank, but as Gould and Sharpe indicated, without acting, they all 
appear to rank with the first-named species of Mirafra javanica. 
Hartert in Nov. Zool., Vol. XII., p. 237,1905, gives the forms of this species. 
In 1912 I criticised a good series of birds, and I could not specifically separate 
the Australian specimens from M. favanica. There are slight differences hi the 
strength of the biU, the shorter minute first primary, the slightly more romided wing 
and the shorter claws, but aH these featmes vary inter se that no good separative 
, Mirafra javanica queenslandica Mathews. 3 
/ 
rnmUMi 
