Are we not authorised to apply this principle to the length of M. Caillié’s stages, since the number of the lines of route is not less than six hundred and thirty three? I may add that the same remark applies to the angular deviation. Indeed, considering the total distance between the meridian of Timé and that of Fez, or the difference in longitude (which is equal to one degree forty four minutes) as being the real sum of all the angles of the route to the east or west of the first, it will follow that the forming of the lines of bearings noted by the traveller (or the calculation of the angles, which is more exact) gives a total variation, so much nearer to the above difference of longitude as the number of bearings is greater; and this number is also six hundred and thirty three. As it has been seen above, I have found the deviation equal to less than six degrees, and it has been easy to divide it over the whole line.
OBSERVATIONS ON THE SITUATION OF TIMBUCTOO, AND PARTICULARLY ON ITS LONGITUDE.
From the determination of the lines which join Timé, Djenné, Timbuctoo, and Fez, and from the latitude of Timbuctoo thus fixed according to the itinerary at seventeen degrees fifty or fifty one minutes,[84] necessarily results a more westerly longitude of this city, than has been hitherto supposed, and even than that which I had formerly admitted, and which brings its position much nearer to the ocean than any geographer has placed it.
Such is the importance of the situation of Timbuctoo that I deem it necessary to insist again on the discussion of the elements; in order, not to establish it with certainty (I am far from believing that I have accomplished this) but to furnish the reader with new means of approaching the truth, as nearly as the state of our knowledge will permit. It is true that a single astronomical observation made by a skilful observer, furnished with good instruments, and worthy of confidence, might overthrow these inductions; but it is equally true that till we are possessed of such an observation, no geographer can fix this main position on a map, without deducing all the reasons which determine him, especially if he dissents from the opinions of his predecessors.
Timbuctoo is placed, on the general map of the travels, in the latitude resulting from the observation of the shadow and the length of the route from Timé to Timbuctoo. A great part of this route, it may be observed, is in a northerly direction, which was far from being conjectured from the course of the river in this quarter. If Mungo Park could have acquainted us with the particulars of his journey beyond Sansanding, we should not have been so long uncertain of this direction, which has uniformly been carried easterly, and that because the situation of Timbuctoo was imagined (as it is still) to be very central in the continent. Park’s map (travels in 1805) places it under the meridian of Paris; Clapperton at 0° 5° west. Rennell at 2° 3°: but the data of which we have been for ten years in possession oblige us to bring it nearer to the ocean; M. Walckenaer has done so, in adopting a longitude from nine to twelve minutes more westerly. I have always proposed to carry this position much farther west, and even placed it four years ago four degrees west.[85] M. Brué has since adopted 3° 34’: perhaps it should have been advanced as far as the sixth degree. For the reasons I have elsewhere given, the route from Timbuctoo to Fez cannot be fixed at more than two English miles an hour: to this rate I have reduced the computation of M. Caillié, who at first estimated the whole route equally at three miles. Now the line of route thus constructed and resting upon Arbate, a well known position, carries Timbuctoo very near the eighteenth degree north; that is to say, from 17° 50’ to 17° 55’; had not this been shortened a little, the city would have been carried to 19°, or 20°, a position far too northerly, and totally inadmissible. Moreover, the line from Timbuctoo to Arbate would be ten degrees too far west; it would also be increasing the declination to 27° instead of 17°, which seems to be its amount for the mean meridian.
But, if the bearing of the line from Timbuctoo to Fez resulting from the itinerary, and that of the line from Timé to Timbuctoo, are preserved, these two intersect each other about the parallel of 17° 50’. This latitude agrees so well with that resulting from the height of the shadow, 17° 51’ that we derive from these data a very satisfactory confirmation. Now the line of the parallel 17° 51’ and the two lines of direction meet, all three, on nearly the same meridian, the sixth west from Paris.
No positive data are yet known which can better fix the longitude: not only are the itineraries of the Moors too vague, but they support themselves upon the Mediterranean, while M. Caillié’s routes, though they are continued to the Mediterranean, support themselves upon the ocean; which is much nearer than the former, and also upon the known positions of Timbo and Fez.
Suppose, however, it were determined, notwithstanding all these reasons to withdraw Timbuctoo farther to the east by a degree or more, it would then become necessary greatly to increase the journeys from Kakondy to Timé. In making them three English miles and three fifths an hour, the true measure has probably been somewhat exceeded, but the situation of Timbo compelled this; while, to place Timbuctoo on the third degree of west longitude, would be to suppose a march of more than four miles and a half an hour.
Another circumstance here comes to our assistance: it is that by carrying it farther towards the east, we must suppose the traveller greatly and constantly deceived in a western direction on his journey from Timé to Timbuctoo, and, on the contrary, in going from Timbuctoo to Fez, he must have made a similar and not less considerable error in all his bearings towards the east. How should we account for this singularity?
From all these considerations, could I risk changing the result of M. Caillié’s itinerary and altering the construction of his route? could I, in short, without sufficient motive, give the preference to any one map over another? Some judgment may be formed upon this question from the diversity of situations assigned to Timbuctoo by different authorities: