356 



which is really bisexual may be taken for a male. Another 

 difficulty is that the anthers fall early and even the filament 

 sometimes disappears from the open perianth, so that the 

 somewhat similar mistake may be made, at a later stage, of 

 considering a bisexual flower as female only. There is, how- 

 ever, considerable irregularity about the sex of flowers and in 

 some cases there appears to be a tendency for the upper flowers 

 of the spike to be male only. Much further work is required 

 in the examination of living specimens. 



The lobes or teeth of the perianth have been described by 

 most authors with greater fullness than in some cases they 

 deserve. The examination of living plants shows that in 

 Arthrocnemum halocnemoides and -4. arbuscula the young 

 perianth completely encloses the male and female organs 

 without any perforation at the summit. Probably the texture 

 is thinner above the stamens and style, and as these develop 

 they push through the perianth, leaving an irregularly 

 lacerated opening. In Arthrocnemum Lylei, on the other 

 hand, each perianth is divided at the truncate summit into 3 

 equal deltoid lobes, and in Pachycornia robusta there are 3 

 or 4 unequal lobes. As regards other species further researches 

 should be made in the living plant. 



The articles are, in the majority of species, so much alike 

 that they afford an uncertain njeans of distinguishing between 

 them. The exceptions are Tecticomia cinerea, Pachycornia 

 robusta, and, to a lesser extent, P. tenuis. The fruiting 

 perianth, pericarp, and seed are a much surer guide, as their 

 characters are strongly differentiated and remain constant 

 within the species or variety. The wrinkling and granulation 

 of the testa in the rough-seeded species seems to be due to a 

 shortening and contraction of the cells towards the back of 

 the seed. 



The fruiting perianth is various in texture, from thin and 

 membranous to thick and spongy. The pericarp varies still 

 more. It may be a delicate, hyaline membrane, often difficult 

 to find, which breaks away from the base of the perianth and 

 remains attached to the upper part of the latter, or it may 

 become hardened and almost horny, or it may, along with the 

 perianth, become more or less absorbed in the enlarged and 

 hardened rhachis ( Pachycornia ). In Arthrocnemum and 

 Salicornia both perianth and pericarp usually open at the 

 base before they fall from the spike, and the seed escapes in 

 this manner. In Tecticomia the perianth splits into 2 

 segments or valves and the seed has already escaped from the 

 base of the delicate pericarp. In Pachycornia the spike doubt- 

 less falls to the ground, and sun and moisture in time split 

 open the bony axis and release the seed. 



