Origin and Development of the Composite. 15 
Bolivia has a number of very dwarfed shrubs, only an inch or two 
high, the majority, however, are erect and from six inches to two 
feet high, as in the other Andine regions. An interesting spiny species 
S. spinosus, D.C., occurs in the Lake Titicaca region, where the 
dryness and excessive insolation (cp. Sect. A above) may be 
judged from the fact that the lake is gradually drying up by 
evaporation although supplied by seven rivers which are of con¬ 
siderable volume during the rainy season (52). 
The shrubby forms of India and China are scandent in open 
forest or at the edges of denser forests as in Brazil ; the erect 
shrubs are more or less confined to the Deccan and the mountains 
above the tree level, as in South America. 
In Africa shrubby erect and scandent forms occur along the 
mountains from Abyssinia to Cape Colony, and a number of trees 
occur at high altitudes; of these none is more striking than 
S. aduivalis, which forms an open forest on Mt. Ruwenzori (14) 
similar to that formed by Espeletia grandijlora on the Paramos 
of Colombia (cp. I, 42, Fig. 109 and Goebel 32, Teil 11, PI. X and 
p. 17). In South Africa the- peculiar shrubby form is more coni¬ 
ferous or “abietoid,” and is to be compared with ericoid species of 
the same latitudes in the Andes. 
The shrubby species of New Zealand are also of a distinct 
and peculiar type, but most of the Australian arborescent species 
are more allied to the South African forms. Each region in fact 
has its characteristic type of shrubby Senecio. 
From the account given in Chapter IX it will be clear that the 
fruit dispersal of Senecio is easy on grassy plains and on mountain 
sides, both above the tree level and in the unwooded regions which 
are common below a certain altitude (cp. 34, p. 243 and I, 7, p. 484). 
The tundra zone of the northern latitudes also holds no obstacle 
to the dispersal of pappose fruits and the wide areas of the Arctic 
species is the result. 
In migrating from its place of origin along the mountain 
ranges of the world Senecio has obviously followed the line o^ 
least resistance and avoided all forests and low-lying regions 
where fruit-dispersal by wind becomes distinctly problematical. 
The diversity of conditions along such a path (cp. 35) as well as 
constitutional instability is probably responsible to a great extent 
for the concentrations of local species which occur at all favourable 
points. 
Conclusions. 
Taking Bolivia as the still hypothetical centre of origin for 
