CHAP, xiii WESTERN WATERSHED OF THE JORDAN 



J55 



It is, therefore, quite possible to account for the existence 

 of the Esdraelon Gap without assuming that it ahvays marked 

 the watershed between the Mediterranean and the Jordan (for 

 this, indeed, probably lay as far west as Cyprus) ; while a con- 

 nection between the Jordan and the Nile would in no way help 

 to explain the difficulty for which it is proposed. Such a 

 connection is directly disproved by the fact that the Medi- 

 terranean fish fauna is absent from the Jordan ; only one or 

 two species, and those are Blennies, belong to both. If there 

 had been any direct connection between the Jordan and the 



Walker Sr Iloutall sc. 



Fig. 15. — The Neighbourhood of the Esdraelon Gap. 



rivers of the Mediterranean basin, it is inconceivable that the 

 fish of the latter should not have entered the former. More- 

 over, any such connection would be useless to our present 

 problem, for it ignores the real difficulty. If we follow the 

 Nile downward from its upper basin, genera of fish such as 

 Clarotes, Lepidosiren, Gyninarchus, and HaplocJiibis disappear, 

 and others not known in the Upper Nile take their place. 

 Thus JMocliococus replaces its close ally RJiinoglanis. In the 

 Lower Nile there appear certain fish, such as five species of 

 Mullet {Miigil), which have entered from the Mediterranean, 

 and another group, which appears to have entered from the 

 equatorial region, at the time when the Zambesi and the 

 southern lakes received their contribution from this source. 



