78 



Banks of Avon River, Beverley (Dr. F. Stoward, No. 270). This. is our furthest 

 east in practically the latitude of Perth. We break away south-east in comparatively 

 dry country (for E. rudis), and find it at Phillips River (Cecil Andrews), which is north- 

 west of Esperance, and more east than ony other locality known to me. 



Now we return to the coast, and it occurs " between the Moore and Murchison 

 Rivers " (E. Pritzel, No. 639). Almost typical E. rudis. 



Large tree on pasture land. Narrow lanceolate leaves. Buds somewhat small 

 and showing commissure. Between Ebbano and Yandanooka, Victoria district (Dr. 

 A. Morrison). 



Then I found it in the bed of the Irwin River, near Dongarra, and as I studied 

 with the view to compare it with E. rostrata, my notes will be found at p. 79. 



Fruits rather small, and the rim domed rather than sharp, as in the normal form, 

 Northampton (J.H.M.). This is about 30 miles north of Geraldton, and is the most 

 northerly locality recorded so far as I know. Diels and Pritzel refer to " indistinct 

 tropical forms " between Northampton and the Murchison, but I have not seen them. 



AFFINITIES. 



1. With E. rostrata Schlecht. 



" It is . . . almost linked by exceptional transit forms with E. rudis, which 



takes its place in littoral South and West Australia. . . ." 



" In E. rudis the bark is extensively persistent and lough, the leaves are often broader, hardly so 

 regularly and distinctly feather-veined, the flowers are fewer in the umbels and mostly larger, the calyces 

 are often dark-coloured, the lid is almost conical, the half-ripe fruit somewhat bell-shaped on account of 

 its prominent narrow slightly expanding margin, the ovary is then more sunk, the ripe fruit is usually 

 larger, less or not rounded at the summit, but rather semiovate, not very convex, nor very wide at the rim, 

 by which means the exserted portion is more evidently shorter than the tube of the fruit-calyx, or the 

 valves may remain even half enclosed ; to these distinctions may be added, that the leaves of young seedling* 

 are roundish and almost sessile, not narrow-lanceolar as in E. rostrata." (" Eucalyptographia " under 

 E. rostrata). 



Diels and Pritzel's remarks (Englers JaJtrb., xxxv, 44-2. 1904) may be translated 



as follows : — 



When Mueller recorded Eucalyptus rostrata from localities in Western Australia going south to the 

 Murchison River, he referred to intermediate forms between that species and E. rudis. 



" We saw E. rostrata (small fruits) with indistinct tropical forms in alluvial muddy localities between 

 Northampton and the River Murchison. Typical E. rudis (large fruits) was only observed in abundance 

 in south-west regions (that is, between the Blackwood and Vasse Rivers). 



"Almost all (rarely 'exceptionally/ as Mueller thought) the forms of the intermediate localities 

 agree in the structure of the fruit closely to typical rostrata, but with E. rudis in size. 



" We obtained it from the Swan, Avon. Irwin and Greenough districts as ' Flooded Gum.' ' 



