In Appendix A the reasons which showed the necessity for a considerable correction to the original observations made since our arrival at Mogador are fully explained. The difference in the resulting heights throughout our journey is not of much moment as regards the lower stations; but it increases rapidly with increasing altitude, and in the case of the Tagherot Pass amounts to about 500 feet. As this correction was disclosed only after careful examination and comparison of all the observations, the first result, which was derived from the rough reduction made at the time, communicated in a letter from Hooker to the late Sir Roderick Murchison,[2] and which appeared in other published notices of our journey, is probably erroneous to the extent above mentioned. Assuming our final results to be pretty nearly correct, and having been unable to hear of any other easy or frequented pass in this part of the range, we seem to be justified in concluding that this section of the Great Atlas chain, as compared with any of the mountain systems of Europe, maintains a remarkably high mean level. The height of the projecting summits in the adjoining portions of the chain was variously estimated by us at 1,500 or 2,000 feet above the Tagherot Pass. Taking the lower of these estimates, and assuming the other depressions to be no higher than Tagherot, we should have for the mean height of the main ridge at least 12,200 feet. Judging from all the distant views we were able to obtain, the portion of the Atlas chain near the head of the Aït Mesan valley is very similar in character to that extending eastward to the sources of the Oued Tessout, and does not reach a higher elevation. If this opinion be well founded, we have in this part of the Great Atlas a range, fully 80 miles in length, which in its mean elevation surpasses any other of equal length in Europe, or in the countries bordering on the Mediterranean. The chain of the Pennine Alps, from the Col de Bonhomme to the Simplon, alone approaches the same limit, as, excluding those limiting passes, the mean elevation of the dividing ridge for a distance of over 90 miles is about 11,800 English feet. That of the Mont Blanc range, from the Cime des Fours to the Pointe d’Orny, probably equals the mean height of the Great Atlas, being about 12,300 feet, but this is only about 25 miles in length. Excluding the mountains of Central Asia, and the Andes of Bolivia and Peru, neither of which can be spoken of as mountain chains in the ordinary sense, the only considerable range surpassing the Atlas in height is the higher part of the Caucasus, between the peaks of Elbruz and Kasbek, whose mean height, for a distance exceeding 110 miles, must reach, if it does not surpass, the limit of 13,000 feet.