79 



26 (2) E. robusta Sin. 



(a) Calyx-tube. 

 (6) Calyx-rim. 



(c) Yalves, almost united. 



(d) Cells. 



Types of dehiscence showing the valves disunited. 



2c (3) E. globulus L'Herit. 

 (a) Calyx-rim. 

 (6) Valves not exsert. 

 (c) Cells. 



2d (4) E. rostrata Schlecht. 



(a) Calyx-tube. 



(b) Disc (capsular). 



(c) Yalves exsert. 



2e Partie sphacelee des valves. 



In species with a long operculum, the valves of the capsule are always conspicuous, and in a few 

 species the valves remain united at the top of the ripe and empty capsule. In all the Cornutse there is a 

 tendency thus to cohere at the tips, especially when young, but the valves usually separate as growth 

 proceeds. There is some shrinkage of the lower portion of each valve, and thus longitudinal apertures 

 are formed, through which the seed is discharged. 



E. cornuta Labill. If we turn to Plate 142, figs, la and lb, and Plate 143, fig. 2a, it will be seen 

 that the upper portions of the valves are not coherent. 



The following specimens have the upper portions coherent, i.e., fused at the tips. 



No. 225, E. Pritzel, Dist. S. W. Plantagenet ; King George's Sound (B. T. Goadby) ; King George's 

 Sound, Lower King River (J.H.M.). Many of the valves are united at the tips, others are free; Cape 

 Naturaliste (J.H.M.) ; Wonnerup, near Busselton (Forest Ranger Donovan). 



In E. Lehman ni Preiss, an allied species, we have tips of valves united in the'following specimens • — 

 Wilson's Inlet (S. W. Jackson); Cape Riche (J. Wellstead); Eucla (correspondent of J. Staer). See also 

 figs. 36 and 5c, Plate 144. 



E. anvvlata Benth. See Plate 145, fig. Id. 



E. Risdoni Hook, f., var. data Benth. 

 3. Juvenile foliage of the type of E. hypericifolia R. Br. " in coll. saxos prope Risdon Cove (Hobart), 

 Robert Brown, 1802." 



x E. Chisholmi Maiden and Blakely, n. sp. 

 4a, mature leaves ; 46, immature buds, with bracts ; 4c, views of not perfectly mature anthers ; id, fruits, 

 not quite mature. On the ridge near the new railway tunnel, Marrangaroo, 5 miles west of 

 Lithgow, New South Wales. (W. F. Blakely and Dr. E. C. Chisholm, 8th May, 1922.) 



E. Gardneri n. sp. 

 5a, juvenile leaf ; 56, 5c, mature leaves ; 5d, 5e, mature leaves ; 5/, umbel of young buds ; 5g, different views 

 of anthers; 5h, calyx-tube showing persistent style; 5i, fruits. All Wagin, Great Southern 

 Railway, Western Australia, except 5c, which is Bendering. (C. A. Gardner.) The type. 



PLATE 255. 



x E. Taylori, n. sp. 

 la, juvenile leaf, as early as has been seen; 16, twig with buds and flowers; lc, different views of anther; 

 Id, twig with fruits. Note that the leaves of Id are narrower and more mature than those of 16. 

 State Forest No. 596, Spring Ridge, Gunnedah District, County of Pottinger, New South Wales. 

 (Assistant Forester T. W. Taylor.) 



