( 237 ) 



THE FORMS OF AHRAFRA JAVANICA. 



A luost: interesting — because of its various subspecies — species of Miiafra, is 

 spread over the Easteru Arcliijielago, from Java to the I'hiliiipitios and Australia. 

 I am acquainted with tlie following forms : 



1. Mirafrajacanwajacanica'iAoviii. 18^U. llah. : Java, Bali. 



'Z. Mirafra javanica parva Swinh. 1.S71. llah.: Flores, Snmliawii, Lombok, 

 Savn, Snmba. 



'i. Mirafra jacanica jiliilijijiincnsis Wardl. Rams. Issti. Hah.: Philippines. 

 Apparently onlj' Luzon. 



4. Mirafra ja.vanica horsfiddi Gould lS47. Hah. : Probably only northern 

 and easteru portions of Australia. We have specimens from the following 

 localities : Moreton Bay, Eurelia, Glencoe, South Alligator River. The upper 

 surface of tliis form is very deep brown, the feathers of the back and scajiulars 

 almost black with pale rufous edges. 



5. Mirafra jamnica pallida Hall, " Emu " 1904. Differs at a glance from the 

 true M. J. horsjieldi by its greyish upper surface. The edges to the feathers are not 

 dark browu or rufous, but ashy grey, the rump and upjier tail-coverts greyish, 

 the under surface cream instead of rufous-buff, the edges to the quills paler. We 

 have two examples from Mr. Hall, one without indication of locality (25. x. 1902), 

 the other said to be from " N.W. Australia," from Rogers, taken 13. xii. 1902. 

 " Iris brown ; bill : upper and tip of lower brown, base of lower yellowish white; 

 corner of mouth pale yellow ; feet and legs very pale brown." These two speci- 

 mens are alike. One bears tlie name ^pallida" in Mr. Hall's liaudwriting, and 

 may have served for the original description just published. It will be the 

 duty of our Australian brother-ornithologists to find out the exact distribution 

 of this new subspecies, and to tell us ou what soil and in what sort of country 

 it lives. 



We have also a male from Swan Hill, Victoria, S. iv. 1899, also from Mr. 

 Robert Hall, which is paler and greyer than typical korsfieldi, very near to 

 yrisescens, but not quite like it. Is this merely an exceptionally pale and grey 

 hofsjieldi, or another new race ? 



6. Mirafra javanica secunda Sharpe 1890. Distinctly more sandy and paler, 

 not so blackisli above, as M. j. horsp'i'ldi. Hah. : Southern and western portions 

 of Australia (Kimberley, Port Headland, Derby). 



7. Mirafra javanica woodwardi Milligan, 190l. Above not at all blackish, 

 browu, or greyish, but bright cinnamon, abdomen ])ale cinnamon buff, throat 

 paler, upper and under wing-coverts briglit cinnamon. Chest-markings not 

 blackish-brown, but ferruginous. Hah. : Sandy jiortions of Onslow and Point 

 Cloates (Tom Carter coll.). 



201. Poephila personata Gould. 



I'uiiilida jicinDiiulu Goulil, I'.Z.S. 1842. p. 18 (" NurthiTii parts of Australia "). 



According to what I ca]i tiiid about the distribution of these birds, this forin 

 and leucotis wonkl occur in the same districts. This is not probable. /-". persoiuita 

 and F. leucotis are most likely subspecies representing each other. The latter 



