23 



After consideration, in I'art XVJI of the present work, I coiustituted E. 

 Drummondii as a variety of E. OklfieUUi as already stated, adopting Drunimond'.s No. 8G 

 (the type of E. Drummondii) as the type for the variety. 1 am jiow of opiniun that 

 E. Drummondii is a vahd species. 



RANGE. 



It is confined to Western Austraha. As in the case of so many other of 

 Drummond's specimens, we do not know precisely their localities, but inasmuch as it 

 lias only been certainly found since from the York district, we have an indication 

 of Drummond's locality, and I would urge systematic search for the species. Local 

 observers are now aware that it has long been confused with E. Lcme-Poolei (a species 

 to which it is more closely related than E. Oldjieldii), and this should faciUtate search. 



Drummond's No. 86. The inflorescence varies in size somewhat in various 

 specimens. Figured at 3 and 6, Plate 74. 



The following specimen matches the type absolutely : — 



Small tree of about 20 feet. Trunk and branches smooth, whitish buff, with a 

 few brown semi-detached scales of dead bark. Leaves dull green. Growing in light, 

 humous soil, mixed mth ironstone gravel. Cut Hill, York (0. H. Sargent, No. 266). 

 (Figured at 5 and 7, Plate 74.) 



Also St. Ronan's Well, near York (C. E. Lane Poole). 



The following specimens have been examined : — 



No. 86 (Drummond). Herb. Cant, and Herb. Oxon. The former in bud (one), 

 but mostly early fruit. The latter mostly in bud and flower, and a httle early fruit. 



AFFINITIES. 



1. With E. Oldfieldii F.v.M. See p. 21. 



2. With E. Lane-PooleiM-Men, in Proc. Roy. Soc. N.S.W. Ini, p. 107, (1919). 

 This is its closest affinity, and will be dealt with when E. Lanc-Poolei i? reached 



