Page:The Journal of Indian Botany.djvu/481

CUEEENT LITEEATUEE. 59 the comparison with Acacia phyllodes as indicating that the simpler types of Irid leaf are petiolar phyllodes. She has further come to the conclusion that the same explanation applies to the more complex types of leaf met with in the family, since these prove on examination to be readily derivable from the simpler types, among which the ensiform leaf is regarded as funda- mental.

The writer considers that leaf-evolution within the family has proceeded on two distinct lines ; on one line the essential feature has been the reduction and ultimately the complete loss of the petiolar region of the phyllode ; on the other line this region, instead of being lost, has been developed and elab- orated.

Moraea Sisyrinchium may be taken as an example cf the cases in which reduction of the petiole plays the chief role. The greater part of the leaf is flat and dorsiventral, but it ends in a short cylindrical apex, interpreted as a vestigial petiole, while the leaf base forms the main part of the leaf.

The second line of progression from the simpler types — that of the elabora- tion of the petiolar region — leads us to the more peculiar leaves among the Iridaceae, some of which are so anomalous that it requires a considerable effort to realise that they are leaves at all. The leaf of Gladiolus tristis, for instance, is cruciform in transverse section, while that of Cypella is of an even stranger type, and looks as if it were elaborately folded. The clue to the peculiarities of both the^e leaves is to be found in the transition from leaf- sheath to limb. In both cases the study of the extreme base of the limb suggests that it is a cylindrical petiolar phyllode, modified, however, by invaginations and keels. It is the extreme point to which this invagination is carried which is the main factor in the production of these aberrant leaf- types.

The writer's examination of the leaves belonging to the two main tribes of the Iridaceae (Iridoideae and Ixioideae) has led her to the conclusion that they can all be explained as phyllodes variously modified from the ensiform type. The third tribe, the Crocoideae, presents at first sight rather more difficulty. But, in this case also, the study of the transition from sheath to limb reveals the fact that we are dealing with a petiolar phyllode. The leaf of Crocus itself is a very peculiar one, and has hitherto been regarded as entirely isolated, and as showing no morphological relation with any other type in the family. But when the Crocus leaf is considered in the light of the Phyllode Theory, it at once falls into line with those of other plants belong- ing to the tribe. The writer interprets it as a petiolar phyllode modified by paired invaginations, and regards its resemblance to an ordinary bifacial blade as purely superficial.

The members of the Iridaceae have all had the same problem to solve — that of how to transform a petiolar phyllode into something with a more adequate assimilating surface — and they have apparently discovered only two methods by which to achieve it, namely invagination and the development of keels and wings. But the varying way in which they have applied these monotonous methods has resulted in the evolution of an astonishing variety of " pseudo-laminae."

Printed and Published for the Proprietor by J. B. BUTTRICK at the Methodist Publishing House, Mount Road, Madras.