BOOBOOK OWL. 
Spiloglaux novceseetandice ocellata (Bonaparte). 
South-west Australia. 
This form is lighter than the three last named and comes closer to the 
typical hoobook from New South Wales. It has the under-surface striped, no 
spotting or marbling being apparent. It is slightly darker than the New 
South Wales form and has been generally called hoobook. As showing the 
variation observed in rare cases, Ashby has written of a specimen secured 
at Bayswater, near Perth : “ The coloration is much darker than any 
specimen of this species I have seen in S(outh) A(ustralian) collections. The 
usual spotting is almost entirely absent, and the tail is not barred.” 
The name ocellata was used by Sharpe for the North-west Austrahan 
bird, but upon examination of the type it was found to agree exactly with 
South-west Australian specimens and to appreciably differ from any (and 
a large series were available) bird from the North-west of Australia or 
Northern Territory. 
Spiloglaux novceseelandice mixta Mathews. 
Parry’s Creek, North-west Australia. 
This is the name to be used for the form crossing North Australia from 
Cloncurry River to the North of West Australia. It is the most pallid form, 
and has been recognised by all workers under the name ocellata,^ as mentioned 
above. It is very pale above and is streaked with pale rufous brown below, 
in some cases the streaking being so light as to give the impression of a 
white under-surface. It varies a little in shade, and it is possible more than 
one subspecies is represented in the stretch of country outlined, but there 
is not sufficient material available to indicate such. 
Spiloglaux novceseelandice melvillensis Mathews. 
Melville Island, Northern Territory. 
This is an island form which must be admitted as it cannot be lumped 
with the preceding, the nearest geographical race, and it appreciably differs 
from all others. It is a small dark race, recsSlmg maculata=clelandi, and 
not at all like the mainland form. I have mentioned that probably 
halmaturina, which I here disallow, is a valid subspecies and this subspecies 
emphasizes this suggestion. I will again note the frequency of constant island 
forms of this species around Australia and also their prevalent darkness. 
Spiloglaux novceseelandice lurida De Vis. 
Mid Queensland. 
I have described and figured the bird I named Spiloglaux boweri, which 
I now conclude to be De Vis’s lurida. I wrote : “ Differs from all sub- 
species of hoobook,^ ^ and this race is a very striking one. It is perhaps a 
331 
