28 ON THE STUDY OF ZOOGEOGRAPHICAL REGIONS, 



needed for general study of this group of rushes should be drawn 

 completely around the whole region in which they occur, and 

 not as a number of small circles around the various waterholes. 

 So, also, in mapping contours for groups of Odotiata, we do not 

 draw our contours along the boundaries of rivers and lakes* 

 though the species are actually confined to them; rather, we 

 include the whole area in which, given, the necessary ivater^ the 

 particular species can be shown to occur. 



B. llie gene7'al theory of Specific Contours. 



Let us select for study a region, Z, separated by a definite 

 barrier from another region, Y. A group of species occurring in 

 Z may either have originated in Z, or they may have immigrated 

 into it from some other region. Suppose a group of species. A, 

 to have been inhabitants of the region Y at some past time before 

 the barrier between Y and Z was effective, and let A be a domi- 

 nant or increasing group. As it extends its boundaries, first one 

 and then another species may reach Z and penetrate further and 

 further into the new region. As these new arrivals encounter 

 new conditions of life, such as altered temperature, rainfall, 

 geological or vegetational conditions, their progress may be 

 gradually stopped. Some forms may penetrate further than 

 others, or may take different paths. As long as the barrier 

 between Y and Z is not a complete one, so long will this immi- 

 gration stream flourish and be clearly recognisable as such. 



The contour of such an immigration gi-oup over the area Z is 

 easily recognisable (Transparency 1 ) by the fact that its zoocentre 

 either lies entirely outside or only partly inside the region Z, 

 while the lower contour-lines extend farther and farther into the 

 region. 



For such a contour, the name EGtoye7iic Contour is proposed. 



Suppose, next, that the barrier between Y and Z becomes 

 complete, so that the immigrant-stream is cut off" from the parent 

 group. If it does not die out, it will gradually assimilate itself 

 to the new conditions, forminy new zoocentres in those areas 

 where co7iditions are most favourable to it. After a suflicient 

 interval of time, it will have evolved a group very distinct from 



