APPENDIX. 
xxvii 
direction, as well as the particular ?iature^ and tendeney^ of the curves, of 
the lines of equal quantities ? 
The variation lines on the globe have occupied a good deal of my attention 
at different periods of my life, and therefore the application of such new 
observations as the assiduity and kindness of my friends had procured for 
me on this occasion, was less difficult, than if the subject had been new to 
me. A dissertation on the subject, would be out of place here ; and there- 
fore I shall only give the result of my inquiries, in abstract; after premising, 
that the theoretical part belonging to the interior of Africa, is founded on a 
supposed conti?iuation of those lines of equal quantities, whose tendency has 
been already ascertained, in the surrounding seas. I am perfectly aware, 
that some may regard the assumption as too great: but they will no doubt, 
admit, at the same time, that it is difficult to conceive . a more probable 
arrangement : and what is much more to the purpose, is, that if we arc 
compelled to abandon the system, in the gross, the quantity of variation in 
the line of Mr. Park's travels, cannot be gready different from what wc 
have assumed. For, whether the line of 18° in the south Atlantic^ be a 
continuation of that in the north Atlantic, or of that in the Indian sea, much- 
the same result will follow : only that in the former case, the quantity will 
be somewhat greater. 
It would appear, that between the East Indies and South America, Europe 
and South Africa, there are four distinct sets of what may be termed con- 
centric curves of variation lines, on the globe, and whose highest points of 
convexity are opposed to each other, within the great body of Northern 
Africa. The accompanying sketch will best explain it.* It would appear 
moreover, that from the place of opposition of these curves, in Africa, where 
* This sketch is not pretended to be minutely accurate ; it being morally impossible 
to procure recent observations in every part, from the rapid change that takes place in 
the quantity of the variation, in one and the same spot. However, the observations that 
determine the course of the lines in the Atlantic (and which are marked on the sketch) 
are from observations so late as 1 79^3. The same is to be said of those in the western quar- 
ter of the Mediterranean ; and those beyond the Cape of Good Hope, to longitude 30* 
east, are of the year 1789. 
It is obvious that a critical knowledge of the quantity of the variation in any particular 
d2 
