The area of South Island is about fifty-eight thousand square miles, approximately equal to that of the state of Illinois, or of England and Wales taken together. Elongated northeast to southwest it extends for some six hundred miles between 40 degrees and 47 degrees south latitude. For most of that distance its width varies between a hundred and a hundred and fifty miles. The coastline (see end papers) is not greatly indented except in the extreme northeast, where subsidence of the land has drowned old river valleys to form the fretwork of bays and channels of the Marlborough Sounds, and in the fiorded strandlines of the southwestern tip. Rough terrain in the immediate hinterlands of these areas has, however, hampered the development of ports, which have appeared instead wherever they were fostered by economic demand, often in very inferior sites. Elsewhere good natural harbors are confined to inlets eroded from the volcanic debris of the Banks Peninsula, Dunedin, and Bluff areas.
The land forms are built of rocks which vary widely in age and type. Many have not been assigned precise chronological pigeonholes because of the absence of convenient fossil date tags. The surface rocks of Stewart Island, most of Fiordland, and western Nelson are much altered sedimentaries of Palaeozoic or greater age intermingled with granitic intrusions. Covering most of Otago and forming a thin core of the principal mountains through Westland, central Nelson, and into the Marlborough Sounds, are the schists. In these, long-continued heat and pressure have reduced the original rocks to a finely foliated mass, which breaks down into thin leaves and in which nearly all clues as to age have been obliterated. Extremely friable, they yet form some of the highest peaks of the Southern Alps where the nature of the rock is well advertized in the jagged crestlines between intersecting glacially eroded basins.
Hardly less uniform in character than the schists, and embracing almost as large an area, are the graywackes, of which the hills of Southland and most of the mountain highlands of Canterbury and Marlborough are built. Sometimes finer than a typical graywacke, sometimes a rough pebbly conglomerate, these rocks are most commonly a coarse gray sandstone. With little evidence of folding or faulting and in the virtual absence of fossils, interpretations of structure and age remain tentative. It is doubtful, however, that the graywackes were deposited later than the Mesozoic era, and they may be older.
On the flanks of the mountains, chiefly on the eastern margins of the great masses of schists and graywackes, are limited remnants of younger covering beds. Tertiary for the most part, that is, under sixty million years of age, these rocks are softer as well as more varied in character than their older neighbors. From them much of the relatively gentle hill-and-downs country has been carved. Blocks of these newer rocks are also present in the faulted mountain area of the northwest. The large accumulations of gravelly debris and silts which are at the surface in most of the true plains and quasi-plains are hardly to be termed rocks at all; these accumulations were deposited during the last brief million years of the glacial or Pleistocene period.
Five major types of land form are identified in the highly generalized diagrammatic map: mountain highland, Fiordland massif, tilted blocks, hills and downs, and plains. Dominating the island's scenery and occupying a greater area than any other type of land form, the mountain highland, built of the schists, graywackes, and older rocks, extends almost throughout its length and, in some places, for more than half its width. The popular geographical cliché "backbone" has comparative validity only if one imagines a sinuous series of bare, knobby, disjointed, and loosely tied vertebrae. The graywacke areas are, in particular, "structureless" in the sense that there is no orderly system of "ranges" or "axes." Where these are described they are usually interfluves of haphazard trend. The greatest heights and the most striking evidence of deep-cutting mountain glaciation are in the stretch of a hundred and thirty miles between two relatively easy passes: Arthur's to the north linking tributaries of the Taramakau and the Waimakariri, and Haast to the south which connects the valley of the river of the same name with that of the southeastflowing Makarora. Here lie the great ice fields from which spectacular valley glaciers cut down through the subtropical rain forest of the west coast almost to the sea. Above the ice, Mount Cook rears its magnificent twelve-thousand-foot peak to dominate the island. However, most of the highland elevations are rather below five thousand feet than above, and passes at between two thousand and five thousand feet are common if not always easily traversed. The highest peaks and, in general, the watershed lie much nearer to the west coast than to the east.
Glacial action has broadened many valleys within the mountain complex. In such a valley Samuel Butler made his homestead in his days of sheep-running on South Island and from it he drew his description of the approach to Erewhon. Many, too, are the lakes--Wakatipu, Wanaka, Hawea, Ohau, Pukaki, Tekapo, Rotoroa, and others--which are legacies of the overdeepening by valley glaciers or of the damming of drainage channels by piles of morainic debris. Some larger basins which are probably structural also exist, notably the Mackenzie country, its floor smoothed and filled by glacial gravels and occupied by a trio of great glacier-fed lakes.
Fiordland and Stewart Island have not been very thoroughly studied, but the rocks appear to be generally ancient and crystalline. The word massif is chosen to describe Fiordland not only because of its geomorphologic connotation of amass of mountains, but also because of its literal French meaning of a solid block of masonry. The mass has been deeply gouged but there is little evidence of recent major folding or faulting movement. Here, in a region where contemporary precipitation often exceeds three hundred inches a year, the Pleistocene ice accumulations must have been enormous. The ice cap built up on an extensive, rolling, plateau-like surface, and the steep-sided fiords of the southwestern coast, resulted from deep ice-cutting followed by a rise in sea level as the major continental ice sheets the world over melted to release water to the oceans. Indeed the lakes on Fiordland's eastern margin are called "fresh-water fiords" by Jobberns.