22 On the Arabian Frontier of Egi/pt. 



monstrate that such a boundary arm of the Nile did actually 

 exist at the time of which the map is supposed to represent 

 the geography, and that it was then the connecting link of a 

 highly important chain of ancient Egyptian frontier cities 

 and fortresses ; among which will be found the places Mr 

 Sharpe proposes to identify with the stations of the Mosaic 

 itinerary. 



As this inquiry involves a question entirely new to ancient 

 geography, it will be necessary to enter upon a very minute 

 consideration of the physical facts from which the proofs we 

 require may be deduced, and of the historical details by 

 which our conclusions appear to be supported. 



The first step in this investigation will be, to study the line 

 of this river's supposed course, in order to ascertain whether 

 the present condition of the valley in which it ran aifords a 

 clue to the physical causes of its decaj'. A correct delinea- 

 tion of the present structure of the land, and its relation to 

 the water-levels, is here the first desideratum. And for this, 

 the invaluable data contained in the " Journal du Nivelle- 

 ment" of the French Commission — designed for the restora- 

 tion of the ancient canal — furnish all the materials we re- 

 quire. 



On the Ancient Frontier Channel of the Nile. 



The section, No. 2, taken along the line of lowest points 

 from Cairo to the end of the Valley of Seven Wells, in which 

 are the Crocodile lakes, delineates the present state of the 

 valley through which this ancient frontier channel had its 

 course. In this section we have, as it were, the country itself 

 before our eyes : Ave have both the means of arriving at a con- 

 clusion, and of judging whether it be supported by fact or 

 not.* 



This section exhibits a double series of levels along the 

 whole line, to shew the relative positions of the land and the 

 water. 



* The survey of this district was effected by Messrs Devilliers, Duchano}', 

 Fevre, and Alibert. For the details of the measurement, &c., vide Descr. de 

 l''*'gyP'^6, vol. xi. ; Extrait du Journal du Nivelloment. To avoid confusion, 

 the leading and characteristic levels alone are indicated in the section. 



