Appendix A. 
Hipparchus, but which that great astronomer had been unable to realize owing 
to the great lack of materials. 
The identity being thus demonstrated of Ptolemy’s two lakes with Lake 
Victoria and the Albert-Albert Edward basin, we now come to the question 
of the Mountains of the Moon. That these uplands, lofty enough to feed the 
reservoirs of the Upper Nile with their snows, are to be placed amid the 
Abyssinian highlands, and more particularly in the mountains of Semen or of 
Gojam, is a view that must be absolutely rejected. To it are opposed the 
northern position of those mountains, the fact that the Abyssinian heights 
stand altogether outside the upper basin of the White Nile, and perhaps also 
the altitude itself which, although in some of its peaks rivalling that of 
Monte Rosa, is still too low to justify Ptolemy’s statement, if, indeed, the 
Geographer intended to speak of perennial snows. 
The Abyssinian 'Mountains being excluded, only two upland groups 
remain that might aspire to the honour of being identified with the Mountains 
of the Moon. These are the Kilimanjaro-Kenia ( 24 ) and the Ruwenzori groups. 
There is, however, a serious objection. Ptolemy (or the Arab interpolation ?) 
expressly states that the longitudinal axis of those mountains is developed in 
the equatorial direction along the parallel of 12° 30' S. The Kilimanjaro- 
Kenia group is, on the contrary, developed in a direction which departs little 
from the meridian, while that of Ruwenzori has a trend nearly from S.S.W. to 
N.N.E. Nevertheless, this at first sight most formidable objection may perhaps 
be overcome, or better, toned down with a consideration of general hydrography. 
We know that, in accordance with their direction relatively to that of the lines 
of water-parting, rivers are normally divided into transversal and longitudinal. 
The first have a trend perpendicular, or nearly so, to the main water-parting- 
line, while the second are parallel, or nearly so, to the same line. To which of 
these two categories belongs the course of the Upper Nile ? If we regard as a 
line of water-parting the undulating ground traversing Unyamweziland, and 
enclosing on the south the Upper Nile basin, and more particularly that of 
Lake Victoria, then the Upper Nile will be classed amongst the transversal 
rivers. If, on the other hand, tve take as main dividing lines either the steep 
eastern scarp of the great African tableland (the watershed towards the Indian 
Ocean), or else the series of heights separating Lake Albert Edward, the 
Semliki valley, and Lake Albert from the Congo basin, then the Upper Nile 
will become a longitudinal river. Now, bearing in mind the decided trend of 
the Nile in the meridian direction, it is highly probable that we should incline 
rather to the first than to the second view, and accordingly place the region 
where the river rises in an upland tract running perpendicularly to its trend, 
298 
