BY EDWIN CHEEL. 211 



pith which give them a jointed appearance readily seen in her- 

 barium specimens. These, I presume, are the " articulations " 

 mentioned by R. Brown. I have examined numbers of plants of 

 this species throughout the Port Jackson district in the living 

 state, and have never found less than six stamens present. These 

 are best seen on a dull damp morning when the flowers may be 

 found to be fully expanded. The seeds are of a brownish colour, 

 and prominently striate, as may be easily seen under an ordinary 

 pocket lens. 



./. prismatocaiyus is a much taller plant than the former species, 

 and usually grows in tufts, hardly ever producing a creeping 

 rhizome. The leaves are much broader and more compressed, 

 appearing almost flat; they are of a yellowish-green colour and 

 are not hollow inside like those of J. holoschcenics, but have three 

 longitudinal partitions of pith extending from base to apex; these 

 are again divided by transverse partitions, but are not so pro- 

 minent as are those of J.holoschcenus. The stamens are only three 

 in number. The seeds are much lighter in colour, appearing 

 almost transparent, and are obscurely striate. 



The internal structure of the leaves appears to me to be a very 

 important character to be observed in the determination of these 

 two species; and may easily be detected by splitting open the 

 leaves. By kind permission of Mr. J. H. Maiden, F.L.S., Director 

 of the Sydney Botanic Gardens, I have examined all the specimens 

 labelled ./. prismatocarpus in the National Herbarium, which are 

 from various parts of Australia and Tasmania. Some of the Tas- 

 manian specimens closely resemble in outward appearance the 

 plants of J. prismatocarpus grown in dry situations in New 

 South Wales. I find that plants of both species, if grown in dry 

 situations, so closely resemble each other as to appear almost 

 identical, " their natural habitat being in swampy ground." But 

 by applying the test of splitting open the leaves it will be found 

 that the hollow-leaved plants (J. holoschcenus) have six stamens, 

 whilst the longitudinally partitioned ones have only three. A 

 further test was made by me by sowing some carefully selected 

 seeds of each of these species in separate pots which received the 



