Flora of Australia. 131 



Eriostemon gracile, R. Grah. (Proteaceae). In Edinb. N. 



Phil. Journ., xvi., 1834, p. 175 = Eriostemon difformis, 



A. Cunn. 



The former name is given as that of a valid species in the 

 Kew Index, iind the Nat. Herbarium possesses specimens from 

 the Grampians, which are indenticaJ. with forms of E. difformis, 

 the glabrous petals and slightly longer flower stalks of the 

 specimens being variable features in E. difformis. The speci- 

 mens do not exactly tally with the description given in Mueller's 

 Plants Indigenous to Victoria, L, 1860, p. 125, but our specimens 

 appear to be authentic. E. gracile is the older name, but to 

 change the current one would be a frivolous interference with 

 established nomenclature. Bentham seems to have entirely 

 overlooked E. gracile, and makes no mention of it in the Flora 

 Australiensis. 



Eriostkmon intermedius. (Proc. Roy. Soc. Vict., 19, 1907, p. 40 

 = E. DESERTi, Pritzel (Fragm. Phytog. Austr. Dec, 

 1905. p. 320). 



The plant was described before Diels' and Pritzel's work was 

 available. Their description is exceedingly condensed, and im- 

 perfect in several respects, but specimens of their plant since 

 received show that the two species are identical, the older name 

 standing. Pritzel seems to have overlooked the fact that the 

 plant is an interesting connecting link, especially as regards the 

 stamens between the Leionema section of " Phebaliuiii " and 

 EriostenK^n proper. The close resemblance to E. Brucei, which 

 misled Mueller, and to which Pritzel attaches undue importance, 

 is mainly external. 



Euphrasia collina, R. Bv. (8yn. E. Jjrowmi, F. v. M., 

 Fragin., v. 88. (Scrophulariaceae). 



There can be no doubt that Mueller wa's correct in placing 

 four of R. Brown's species in one, but as was pointed out by 

 Bentham, the proper course was to extend one of them to 

 include the others, and so avoid a new name. Even considered 

 as varieties, the line of demarcation is not distinct in all cases, 

 and the type forms show a regular gradation from dwarf, small- 



