xxvia 
APPEN DI X. 
the quantity of variation is 18°, it decreases with great rapidity, and finally to 
nothing, in going eastwards to India, or south-wesiward to south America :. 
a'nd that from the same point, it increases, in going N N towards Ireland, 
or to the opposite quarter, towards the coast of Caffraria. But the change is 
not in any proportion so rapid in the increase, in going north or south, as 
in the decrease, in going east or west. Such are the outlines of this system ; 
by which, if a person was to set out from Cape Verd, or Cape Blanco, to tra- 
verse Africa from W S W to E N E, to Upper or Lower Egypt, he would 
find at setting out, from 15^ to 16^ degi'ees of westerly variation, which 
would increase to 1 8° about the centre of the continent, and afterwards de- 
crease to a less quantity than at his setting out. Mr. Park's travels being 
comprized between Cape Verd and the centre of Africa, will therefore be in 
the quarter that has from i6 to 18 degrees; and the part more particularly, 
where, from the want of observations of latitude, we must depend on his 
compass bearings, nearer iS'' than 17°.* I shall now proceed to the detail 
of Mr. Park's materials. 
Leaving Pisania, Mr. Park proceeded eastward to Medinajt the capital 
place, and at a given time, is of less importance to the present question, than that of 
the bearing of the lines of equal quantities, at any recent period : and this object is, I 
think, tolerably well obtained, by the materials before me. Any change that may have 
taken place since 1793, is in favour of a greater quantity of variation, within the limits 
of Mr. Park's travels. 
In the Atlantic, the increase appears to be about a degree in seven years. In the 
Indian ocean, less : and in the Red Sea, there seems to have been little alteration be- 
tween 1762, and 1776. 
The lines, as far as they are founded on authority, are continuous ; but broken in the 
parts assumed. 
* It is in proof of the existence of this quantity of variation, that, on closing Mr. 
Park's route at Woolli, there appeared to be only a small deficiency of distance on the 
Map, when i7'> variation were allowed. Had the quantity allowed been less, this ought 
not to have happened, as Jarra is placed according to its latitude. 
f Major Houghton's point of outset in 1791 was Medina. His route falls into that 
of Mr. Park at several points, and finally branches off from it about 30 miles short of 
the Faleme river, which the Major crossed at Calcullo, n*ar 20 miles higher up thap 
Naye, where Mr. Park crossed it. 
