The linear direction of a mountain chain, the axis of elevation as we might say, (so sharply hinted at in the very word mountain-chain,) brings out relations which vary not only according to the longitudinal direction itself, but to the lateral extent, the number of mountains, the situation, and the ramification of the chain. If the direction be a straight one, we can rightly speak of an axis of elevation. According to Humboldt’s measurements, this axis in the Pyrenees is 230 miles in length; in the Alps, from Mont Blanc to the Hungarian frontier, 515 miles; the Ural Mountains, 550 to 2042 miles; the Scandinavian Mountains, 1100 miles; the Altai Mountains, 9900 miles; the Kuenlun, 1600 miles; the Thian-Shan, in Inner China, 1700 to 2150 miles; the Himalayas, 1600 miles; the Yablonoi Chrabet, 550 miles; the Aldan, 400 miles; the Ghauts, 760 miles; the Andes of South America, 4400 miles; and the whole Cordillera of North America, 9200 miles. There is often much doubt about the true beginning and ending of a mountain chain, and judgments differ according as they rest on the fact of elevation or on the geological traces of upheaval where they begin to be manifest. Geographers are not agreed, for example, whether the Ural Mountains continue as far north as Nova Zembla, and whether one or two chains in America are to be spoken of as traversing the plateau of Mexico. If there are parallel ranges, it is correct to speak of a transverse axis, running at right angles with the main axis. There is, it is apparent, a marked difference between simple chains and the accumulated parallel chains, where breadth is a prominent element, as in the Vosges, the Black Forest Mountains, the Fichtel range, the Hartz, the Ardennes. The parallel rows form a mountain system. Yet all great chains are made up of smaller ones, of groups at least, and so are mountain systems. Often the grouping is seemingly irregular, a lawless aggregation, but only because our knowledge is incomplete, and the law of arrangement concealed from us. This law is traced in the very geological qualities of the chain, not in the later form. The outer form is often very deceptive, the very convulsions which indicate the surer signs having served to obliterate what we should suppose the most prominent marks. The present of mountains must often be studied in the light of their past. Orography must be interpreted by geology. But the geological surveys of the earth are as yet very imperfect; the outer form has often to be accepted as the only guide. Orography and geology are two sciences which now go on hand in hand.