26 Plowman . — The Comparative Anatomy and 
Stenophyllus , Fimbristylis , and Eriophorum seem to lie more directly 
in the line of descent of Scirpus. Eleocharis is apparently a long-estab- 
lished and much reduced limicolous genus from near the common origin of 
the Order. Dulichium is probably a highly specialized but less successful 
early variant from the central line of descent. 
From what has already been said it will appear that the Cyperaceae 
are a relatively primitive group among the Monocotyledons, and that they 
have suffered a considerable morphological reduction and specialization in 
acquiring their present amphibious to xerophytic mode of life. Their 
closest affinities appear to be with the Juncaceae on the one hand, and with 
the Gramineae on the other. Upon embryological evidence Van Tieghem 
(56) concludes that the Cyperaceae serve as the connecting link between 
the true Monocotyledons and the so-called anisocotyledonous Gramineae. 
This view, however, is not supported by the facts observed by Miss Sargant 
(43) in her extensive studies of the cotyledon of the Monocotyledons. At 
all events, anatomical evidence points to a somewhat intimate relationship 
between the two groups. Nevertheless, they are clearly marked off from 
each other by morphological differences in the structure of the culm, the 
arrangement and insertion of leaves, the character of floral parts, and in the 
pronounced difference in character and occurrence of amphivasal bundles. 
In the work on cotyledonary development already cited, Miss Sar- 
gant (43) ably supports the view that the Monocotyledons are derived 
from the Dicotyledons, or that both spring from a common Angiospermous 
ancestry. The evidence afforded by the developmental and morphological 
data accumulated by recent investigators renders the correctness of this 
view highly probable. We may now add to this evidence the facts learned 
from our study of the Cyperaceae. (i) It is to be recalled that in the 
young condition of the seedling of this group the central cylinder is typically 
tubular as in the Dicotyledons. ( 2 ) The same tubular type of central 
cylinder is of universal occurrence in the floral axis of the Cyperaceae. 
( 3 ) It is only where large and numerous leaf-trace bundles enter the 
central cylinder that we find the typical Monocotyledonous type and 
arrangement of stelar elements. Hence we may conclude that the type 
of central cylinder found in the conservative seedling and floral axis is 
ancestral or palingenetic, while the medullary and amphivasal bundles 
occurring elsewhere are more recently acquired or caenogenetic features. 
But those stelar characters which we find to be palingenetic in the Mono- 
cotyledonous axis are precisely those which are observed to be characteristic 
of the stelar system of the Dicotyledons. In so far, then, as the anatomical 
characters of the Cyperaceae are of importance, they point somewhat 
strongly to a common ancestry of the two grand divisions of Angiospermous 
plants. 
