March 22, 1883] 
NATURE 
ta ) 
493 
like the dicotyledonous habit acquired, the leaves become 
more like those of the opposite class. Thus the Arums, 
with their very unlilylike mode of growth, and their long 
petioles rising high into the open air, have usually a very 
distinct broad lamina, and have the veins accordingly 
branched or netted, almost as in dicotyledons. Very much 
the same type recurs under similar circumstances in 
Sagittaria sagittifolia (Fig. 27). Still more markedly 
dicotyledonous-looking are the leaves of certain very 
aberrant Amaryllids, such as Zamus and the other Diosco- 
ridez, which have taken to climbing, and have therefore 
acquired broader leaves with netted veins between the 
Fic. 20.—Developmen tof Monocoty!edonous stem. 
ribs. Compare with these the like result in Sylax; 
and then look at both side by side with such dicotyledons 
as Convolvulus. The influence of the ancestral type is 
here seen in the arrangement of the main ribs; the in- 
fluence of environment is shown both in the approxima- 
tion of general shape, and in the netting of the minor veins, 
Once more, the ovate type of Listera leads on readily 
enough to the whorled leaves of Paris and Trillium, 
where the venation has become similarly netted. A 
bushy type, like Azscws, develops broad leaf-like ped- 
uncles, which closely simulate the true leaves of dicotyle- 
donous bushes with like habit, such as box or privet. 
Fic, 21.—Development of Dicotyledonous stem. 
But the widest departure of all from the central mono- 
cotyledonous type is found in leaves like those of the 
tropical arborescent forms—the palms, screw-pines, &c. 
Most of these have long pinnate foliage, whose origin 
may best be considered when we come to examine the 
Ry ) xf 
: 
Fic. 22.—Development of Monocotyledonous seedling. 
( 
chief dicotyledonous types ; meanwhile such forms as the 
cocoanut or the date-palm may be advantageously com- 
pared, as to conditions and general shape, with the tree- 
ferns in one direction, and the cycads in another. The 
bananas cast much analogous light upon the origin of 
these tropical pinnate forms. Where the plant is less 
arborescent, as in Chamerops, the leaf assumes rather a 
fan-shaped than a pinnate development. 
Among dicotyledons it may be fairly assumed that the 
earliest form of leaf was simple, ovate, and nearly ribless, 
or with faint digitate venation. This is shown both by 
the nature of the earliest leaves in most seedlings, and 
the constant recurrence to such a type wherever circum- 
stances are favourable for its reproduction. Hence, as a 
whole, digitate venation seems the commonest in most 
humble dicotyledons ; and the only problem is how 
pinnate venation came to be substituted for it in certain 
cases. The answer seems to be that wherever circum- 
stances have caused leaves to lengthen faster than they 
broadened, and so to assume a lanceolate rather than an 
ovate shape, the tendency has been for the main ribs to 
be given off, not from the same point, but a little in front 
of one another. If the technical botanists will pardon 
such a phrase, the internodes of the midrib, usually sup- 
pressed, seem here to have been fully developed. Figs. 
28, 29, and 30 show the stages by which such a change 
