386 E. C. ANDREWS. 
form endemic sub-tribes, and their mutual relations are 
to be sought in the tropics rather than in some sunken 
cistropical land block, which may be supposed to have con- 
nected the two countries directly during a previous period. 
The Genisteze of the northern hemisphere present an 
appearance quite different from those of South Africa and 
their stamens differ considerably from the Australian types. 
Thus in the northern hemisphere the stamens are either 
monadelphous or diadelphous, whereas in Australia the 
sheath is generally open along the upper side. 
The gorse, broom, and allied forms, of the north are 
xerophytes and often leafless. As with allied types which 
adopt devices to reduce transpiration, they are aggressive 
colonists in waste places in temperate regions. 
After the great shrinkage of the vast Mediterranean Sea 
of the Cretaceous, HKocene, and Miocene periods, the main 
Hurasian types appear to have invaded Northern Africa. 
Genisteze also appears to have reached North America by 
the same route as Thermopsis, either during the late Ter- 
tiary or the Glacial Period. Among others, Lupinus passed 
southwards to Bolivia along the high plateaus extending 
from British Columbia. 
Adenocarpus, Ulex, and Cytisus, may be either of younger 
age or of less vigorous nature than Lupinus. 
Anarthophyllum and Sellocharis represent modifications 
of tropical Genisteze to meet the South American conditions. 
In brief, the development of the temperate forms of 
Genisteze suggests a development parallel to that of 
Podalyriez, each being due to a change of climate of world- 
wide application whereby the widely-diffused and uniform 
original forms of the tropics became modified in extra- 
tropical regions to secondary types, each unconnected land 
block showing development along different lines, the rela- 
