Phytogeny of the Cyperaceae . 25 
gested by Jeffrey (30) and others, to the effect that the Monocotyledonous 
prototype was aquatic in habit, of soft, loose texture, and provided with 
large leaves. Moreover, these primitive Monocotyledons were in all 
probability fresh-water forms, for otherwise they must have taken on 
a xerophytic structure, as a result of which they no doubt would have been 
preserved somewhere in the geological record. Indeed, we may venture 
the opinion that this is precisely what did happen in the course of time to 
the primitive Cyperaceae. Driven by torrential currents or the drying-up 
of streams, these ancestral sedges became adapted to the dry banks, or 
were carried into humus bogs and salt marshes. By the first course 
we might very readily derive the tuberous Cyperus ; by the second, the 
long rhizome of Carex. By either course we should expect the plant 
rapidly to lose its hydrophytic character, with reduction of fundamental 
tissue and consequent decrease in size of members. At the same time, the 
more conservative fibro-vascular system would respond much less quickly 
to this reduction process, and the stelar structure would grow more and 
more compact, with all the attendant phenomena of nodal complexes, 
amphivasal bundles, medullary strands, and cortical leaf-traces. 
Apparently such an extreme degree of specialization and comparative 
isolation as that shown by the genus Carex could have been reached only 
through a very long period of adaptation. Hence we may consider Carex 
to be a highly-developed, relatively modern genus, derived from a successful 
early variant from the primitive Cyperaceous stock. Cyperus , with few 
exceptions, presents the tuberous, scapose habit and a high degree of 
uniformity in structural details. While Carex and Cyperus are thus seen to 
be very clearly defined and somewhat divergent genera, each presenting 
a fairly striking degree of structural unity, it appears that Scirpus presents 
the opposite extreme of great variability. It alone of the entire Order 
refuses to fit as a whole into any scheme of morphological or anatomical 
classification. In Scirpus we find both Chlorocyperaceae and Sclero- 
cyperaceae ; both Amphivasae and Centrivasae ; both scapose and jointed 
culms. Altogether the genus appears to be anatomically in the ‘ experi- 
mental stage ’ ; the focus of evolutionary activity in the Order ; the terminal 
bud, so to speak, of this particular phylogenetic branch. These facts, how- 
ever, do not necessarily indicate that the genus is a recently established one, 
but that it may be a somewhat ancient and generalized type. 
As for the more intimate phylogenetic relations of the various genera 
of the Cyperaceae, it must be conceded that the whole matter is at best 
highly problematic, and we shall venture to add only the most general 
statements in regard to it. On anatomical grounds it seems reasonable 
to place Kyllinga , Dichromena , Fuirena , Hemicarpha , and Lipocarpha on 
the side of Cyperus , though probably of more recent derivation ; Rhyncho- 
spora , Cladium , and Scleria tend strongly toward Carex ; while Psilocarya , 
