38 DR. M. T. MASTERS ON THE 
the cotyledons, show a simpler and less differentiated structure 
than in the leaves. There is no hypoderm and there are no 
palisade-cells. The resin-canals have the same relative position 
as in the adult leaves. It is noteworthy that the structure of 
the primordial leaves of Pinus presents a close approximation to 
that of the 4-sided leaves of Picea, but these have well-marked 
hypoderm. 
The structure of the adult leaf varies slightly in different species. 
Usually there is a well-developed hypo- or exoderm, sinuous 
parenchyma, and a single or divided fibro-vascular bundle ac- 
cording to the shape of the leaf. The resin-canals vary in position, 
being in some species close to the epiderm, in some close to the 
endoderm, and in other species in the parenchyma between the 
two. The cells of the parenchyma are usually sinuous in outline, 
owing to the intrusion of folds of the cell-wallinto the cavity of 
the cell. 
Resin-canals exist in the old wood as well as in the cortical 
parenchyma. 
The structure of the cone-scale with its double fibro-vascular 
layer is well known. The wing of the seed consists of wholly 
regular rows of cubical cells with undulating outline. 
The species occur in the northern hemispheres of both conti- 
nents on the mountains, in the valleys, and by the sea-coast. 
Some extend even into the Tropics, the West Indies, Malaya, 
and the Philippines. The genus was extensively represented in 
the Miocene period. 
GEOGRAPHICAL DISTRIBUTION. 
In the accompanying tables (pp. 40, 41) I have endeavoured to 
show the main features of the geographical distribution of the 
genera. The areas I. to VI. are those proposed by Sclater and 
Wallace and adopted more or Jess fully by most modern writers. 
The 23 “tabulation” subareas(p.42) are taken from Clarke’s paper 
in the ‘Philosophical Transactions’; but instead of arranging 
them in linear series I have placed them, first of all, in a 
Northern and in a Southern division, and then side by side from 
West to East, as they would be in a map on Mercator’s projection. 
In arctic and in subarctic America a very few species extend 
across the continent, but for the most part the species to the 
west and those to the east of the Rocky Mountains, though 
