484 Agues Arber .— The Phyllode Theory of 
succeeded by an upper region flattened in the vertical plane and isobilateral 
in its anatomy ; the phyllotaxy is distichous. The appearance of such 
a leaf at first glance suggests the idea that it has been folded along the mid¬ 
rib and that the two halves of the upper surface have fused ; indeed, the 
opinion that the peculiarities of this type of leaf are due to congenital fusion 
is, apparently, widely held. 1 The actual evidence for this view is, however, 
somewhat slender: neither ontogeny nor comparative morphology afford it 
much support. The leaf, according to Goebel, 2 develops by means of two 
growing points, one belonging to the sheath and the other to the ‘ lamina \ 
The ‘lamina 3 ,except where it passes into the sheath, is solid from the first. 
Goebel shows that the development of the leaf is similar in all essentials to 
that of the ‘ radial } leaves of Juncus and Allium , and he concludes that 
there is, in the case of Iris, no ontogenetic evidence for the concrescence of 
two surfaces. Fifty years previously, a similar conclusion had been reached 
by Trecul, 3 who wrote : ‘ II n’y a point ici de soudure; la feuille nait telle 
que nous la connaissons.’ 
It seems to the present writer that the chief obstacle to a rational 
interpretation of the Iris leaf has been the fact that it is too often considered 
as a case apart and unparalleled. This notion has been fostered by its 
treatment in many elementary botanical classes as if it were a unique 
type. But isobilateral equitant leaves occur in the Iridaceae, and also 
in the Araceae, Liliaceae, Haemodoraceae, Amaryllidaceae, Restiaceae, 4 
Philydraceae, Xyridaceae, and Orchidaceae. It is also perhaps significant 
that the three central Families, Liliaceae, Iridaceae, and Amaryllidaceae, 
contain, in addition to genera with isobilateral equitant leaves, other genera 
with non-equitant leaves, showing the type of structure here called phyllodic. 
The present writer wishes to put forward, as an alternative to the ‘ congenital 
concrescence ’ theory, the view that the isobilateral equitant leaf is merely 
a phyllode flattened in the vertical plane , and thus comparable with the great 
majority of Acacia phyllodes . 
The best justification for this view—which brings the isobilateral 
equitant leaf into close relation with the other Monocotyledonous leaves 
exhibiting phyllodic anatomy—is to be found in a comparative study of the 
Iridaceae. 
Among the Iridaceae there are, besides the vertical equitant leaves, 
others, such as that of Sisyrinchium (Fig. 16 A and B, p. 483), in which the 
upper region is almost radially symmetrical. Such a leaf is closely 
comparable with the phyllodic leaf of Allium Cepa (Fig. 9, p. 479), and with 
the phyllodes of Acacia scirpifolia (Fig. 1 B, p. 474) and of Rhyticarpus 
dijformisl We need only imagine the leaf of Sisyrinchium flattened and 
1 Chodat, R., and Balicka-Iwanowska, G. (1892); Massart, G. (1894). 
2 Goebel, K. (1905). 3 Trecul, A. (1853). 
4 Anarthria , according to Brown, R. (1810), 5 Briquet, J. (1897). 
