138 Alfred J. Ewart: 



term " species " should be given as broad a scope as possible, 

 not merely for reasons of practical utility, though these are of 

 value, but because it is precisely in such genera that groups of 

 varieties as yet imperfectly segregated into species are most 

 likely to occur, and by recognising such varieties as species too 

 hastily we render it more difficult for the workers of subse- 

 quent centuries to obtain evidence of evolution in such cases. In 

 addition, the synonymy is less likely to become so extensive as 

 at present. Thus it is doubtful whether the genus Pultenaea 

 contains many more than 60 valid species, although over 150 

 have been recorded, and the same proportion holds for most 

 large genera. 



A specimen of P. Williamsoni, Maiden^ was referred at Kew to 

 a vaa-iety of P. -stricta. The National Herbarium contains both 

 under P. stricta and P. paleacea specimens examined by 

 Bentham or by Mueller, which come very close to, or prac- 

 tically match specimens of P. Williamsoni. Altogether there 

 can be no doubt that the genus will not be on a satisfactory basis 

 until cultural experiments under varying conditions have been 

 performed with all its supposed species, and tte result of such 

 experiments will probably be to give the tfelected species in the 

 genus a much wider range than they have hitherto enjoyed. 



Sporobolus Benthami, Bailey = S. vihginicus, Kunth, var. 

 PALLIDA. (Gramineae). Queensland Flora, p. 1880, Bull. 

 Dept. Agric, Queensland, xiii., p. 16. 



The 22 sheets of this variety in the National Herbarium show 

 a far greajter range of variation than Bailey's specimen from the 

 type, and yet have no constant character of more than varietal 

 significance. Although the (inter glumes are u>ually about 

 e<:;[ual, the lower one iis occasionalh' slightly longer than the 

 upper, and sometimes, ospeciallj' on the basal spikelets, not more 

 than half its length. In this respect, in the more hyaline outer 

 glumes, and in the longer spike the variety shows am approach 

 to S. indicus, K.Br., from which, however, the vegetative habit 

 differs. It is, in fact, possible that cultm-al experiments might 

 show S. virginicus to be a marsh and maritime foini of variety 

 of S. indicus, developed in brackish situations. 



1 Vict. Nat., vol. xxii,, p. C, 100,=i. 



