404 NATURAL HISTORY OF PLANTS. 
a quite peculiar organization. It is known (chiefly through the 
researches of H. Mout) that in Proteacee generally the stomates are 
very small, and are situated, not at the surface of the epidermis, but: 
at the bottom of a sort of pouch or well, as deep as the epidermis is 
thick, and with the circular or elliptical external opening sensibly 
contracted. The nervation of the leaves is also often characteristic. 
It is pinnate, rarely palmate; the secondary ribs radiate sometimes 
from the base of the blade, sometimes at a certain distance up, like 
the rays of a fan. The terminal veinlets are usually arranged in an 
elegant delicate network, sometimes very complicated.’ The leaves 
often degenerate near the flowers into involucrant bracts, which 
become more and more coloured and simple in form, and recall, in 
their tint and approximation and in the way they protect the 
flowers, the leaves of the involucre of Composite and some allied 
types.” 
Arrinitizs.—The order Proteacee, placed by A. L. pz Jussrev in 
Apetale, was left there by all authors until A. Broneniarz,’ fusing 
this class with Polypetale, placed Proteinee between Rhamnoidee 
and Daphnoidea,—i.e., next the three classes which he terms 
Myrtoidee, Rosinee, and Leguminose. Linpixy‘ puts Proteacee in 
his Alliance XLI (Daphuales), just before Rosales which includes 
Rosacea, Pomacee, Drupacee, Fabaceae, and Chrysobalanacee. In the 
1 It is from these characters that authors have 
thought they could distinguish leaves of Prota- 
ceous plants in geological strata (see Errinas- 
HAUSEN, Proteac. der Vorwelt). Hence a 
detailed study has been made of the nervation, 
which is thus described by Dr Saporta (in Ann. 
Se. Nat., sér. 4, xvii. 248) : “The tertiary ribs, 
always more or less oblique to the secondary, 
ramify by bifurcating to their last subdivisions ; 
the network resulting from the intersection of 
the ramifying venules gives rise to rhomboidal, 
trapeziform., or hexagonal meshes, that vary in 
size, proportion, and regularity with the genus 
and species. These tertiary veins, oblique to the 
secondary, are more or less so according to the 
obtuseness of the angle which the latter make 
with the midrib.” Hence the leaves are divided 
into those which have oblique ribs (Grevillea, 
Lomatia, Leucospermum, &c.), and those in 
which the secondary ribs are given off at an 
obtuse or nearly a right angle (Xylomelum, 
Enightia, Banksia), These considerations have 
led geologists to admit fossil types of Proteaceae, 
such as Leucadendrites, Banksites, Paleoden- 
dron, Lomatites, Knightites, Myricophyllum, 
Rhopalospermites (Sav.), Embothrites, Drian- 
droides (UNG.), as well as true Grevilleas and 
Hakeas. Proteacee are given as “the most 
ancient dicotyledonous type [excluding, of course, 
Conifere] whose presence it is possible to ascer- 
tain in the fossil state.” It is in the Senonian 
beds of the Aachensandsteim [Sandstone of Aix- 
la-Chapelle] that the preponderance of these 
types is most marked, some hundred species 
being admitted. Later on, it is said, in the 
tertiary beds, true Dryandras are found, and 
then the Proteacee commence to diminish in 
number, and seem to be replaced by Myricacee 
(see Sap., op. cit., 298; xix. 21, 58, 109; sér. 
5, ili. 19, 24, 30, 33, 55, 59, 95, 144). 
2 It is more on account of the coloured invo- 
lucres than the flowers that Proteacee are so 
ornamental in the conservatory or winter-garden. 
3 Enum. des Genr. de Pl, Cult, (1843), 120. 
4 Veg. Kingd., 529. 
