in Maranhão, which led nevertheless to the colonisation of the connecting link, Ceará.

Our peripheral colonisation had, of course, great economic consequences. The climatic conditions of the coast compelled the settlers to cultivate colonial products. Sugar was the first of these in which the colonists were interested. The sugar cane was brought to Brazil during the first half of the sixteenth century, and Pernambuco, São Vicente, and later on Bahia were the first areas planted. Cattle-breeding, tobacco, Brazil wood were mere complementary industries. This was the Sugar Period of our commercial history, and the importance of our sugar exports was felt in the international trade of Portugal.

The historical meaning of our coastal topography has had a farreaching influence on our civilisation. Even now, our great centres of expansion and irradiation are on the coast, or within a few miles from it; their history has deep roots in our past, and our peripheral colonisation has been replaced, to a certain extent, by a peripheral civilisation, although new foci of irradiating civilisation are already in an advanced stage of formation on our highlands, chiefly in the southern parts of the country.

II. THE APPROACH TO THE HIGHLANDS.

The second geographical phase of Brazilian history is characterised by the approach to the highlands and by the conquest of the inland mineral treasures, leading us to the Gold Period of Brazilian commercial history.

In South America there are three distinct mountain systems, belonging to different geological ages: -the Andean Cordillera, with which Brazilian topography is not concerned; the Guiana massif, the southern part of which belongs to Brazilian territory; the Brazilian group or highland, which belongs entirely to Brazil.

The latter system is far the more interesting from the Brazilian point of view. The whole country owes to the group its topography and its most prominent social and economic features, for the importance of the Brazilian highlands in the distribution of climatological factors is decisive.

It is a common saying among us that our altitudes redeem our latitudes; one could as well say that Brazil is a gift of the highlands, as that Egypt is a gift of the Nile, or that the British Isles are a gift of the Gulf Stream.

The Brazilian massif, like the Guiana massif, is composed of a core of ancient rocks, overlaid by newer sandstone. The ancient rocks which form the basis of the vast Brazilian upland belong to two principal geological groups. The first, and older, is formed of gneiss, granite, and mica schist; metalliferous deposits are scarce, but coloured gems are to be found. The second group is chiefly composed of schist, quartz, itabirites and limestone; it contains rich deposits of gold, iron, lead, etc. Rocks of this group have been mined in the mountains of the states of Minas Geraes, the Espinhaço, the Canastra and the Matta da Corda.

In the south of the Brazilian highlands, Devonian and Carboniferous beds occupy a vast area in S. Paulo, Parana, Santa Catharina, and Rio Grande.

The coastal range extends from the S. Francisco river on the north to the south of the state of Rio Grande. It belongs to the Brazilian massif, although several parts are broken by wide valleys, like the Parahyba river. The highest summits belong to the Serra da Mantiqueira, where the Itatiaya reaches 10,000 feet. North and South, the great central massif is flanked by lowlands: the Paraguayan plain and the Amazonian plain. These two low-lying areas consist of great river basins, covered by alluvial soil brought from the surrounding orographic systems. The streams have a very slight slope, and both are the highways towards the heart of South America.

About one half of the Brazilian central highland is absolutely unknown, although its principal rivers, the Aroguaya, the Xingú, and the Tapajoz have been explored several times. Lately, the Roosevelt-Rodon expedition has discovered and explored the Rio da Duvida, the largest affluent of the Madeira river.

There is a marked contrast between the northern part of the Brazilian system and its southern part. I wish to draw your special attention to five topographical facts, which explain a great deal of our political and economic history.

First. The northern system, and especially the Chapada Diamantina in Bahia and the Borborema in the north-east, is much farther from the sea-shore than the southern system, and therefore, more ground is yielded for cultivable plains than by the Serra do Mar, which practically slopes into the ocean in the south of Brazil. In consequence the fall-line is much more remote.

Second. The average altitudes of the southern system are far more pronounced than of the northern one. The Serra do Mar and especially the Mantiqueira include the highest Brazilian summits at the meeting point of the states of S. Paulo, Rio and Minas Geraes. The passes are very few, much higher and therefore more important than in the north.

Thirdly. The direction of the southern system is almost parallel to the shore line, and, keeping pretty close to it, allows very few rivers of any importance to reach the ocean, whereas in the north, the directions are different, but generally coastwise, so that the valleys are the real highways towards the interior inland. This fact seems to be decisive in Brazilian history in the seventeenth century. Indeed, the southern system with its few passes constitutes a barrier, whilst the northern system is easily accessible and forms a natural road towards the heights.

Fourthly. The general aspect of the tableland in the north, its peculiar shape, a low, bushy vegetation, a different geological constitution, with different riches and resources, constitute an altogether different environment and emphasise the social and economic contrasts between north and south.

Fifth. The climate, which to a certain extent is a consequence of the topography, has more than anything else accentuated the differences and the contrasts; for the north receives the trade winds under quite peculiar circumstances: the almost rainless district of the northeast has deeply influenced the social and economic activities of its populations.

Many more contrasts could, of course, be found, but the preceding data, although not complete, are sufficient to explain the chief historical

contrasts.

As I said, particularism characterises Brazilian history, and every centre, Pernambuco, Bahia, Rio, S. Paulo, has its own historical development and political meaning, but two cycles could perhaps be given for the sake of simplicity: the southern cycle with S. Paulo as its centre, and the northern cycle with Pernambuco as a type.

Let us note how these topographical facts have influenced the different historical facts of our past. Hundreds of Tupi tribes were found by the newcomers. In the south they were met on friendly terms, in the north they were repulsed by the whites. In the south the Indians were protected by a strong barrier of mountains, and the Portuguese had to come to terms with them, and after the adventurer João Ramalho married an Indian chief's daughter, they were allowed to climb up the highlands, where they were received as guests until their settlements were strong enough to overpower any resistance. In the north, the intention of subduing the natives was evident from the beginning, because it was much easier to get at them.

During that first period of commercial history the sugar cane had been introduced in the northern and southern captaincies, but the topography of the north allowed at once the establishment of a great complementary industry, cattle-breeding, so that very soon the colonists were able to draw their own supplies from the accessible interior, and to feed the new centres not only of the north, Pernambuco and Bahia, but even of the south.

The result was that, so far as home production was concerned, the north became cattle-breeder and the south agriculturist. That very fact explains why during the sixteenth and the seventeenth centuries we find the north-east of Brazil the colonial area of great estates, whilst the south represents the country of small landed proprietors, tilled by the colonist himself without slaves or rural workmen.

The second part of the sixteenth century was devoted to the defence of the new colony. The feudal system of captaincies had proved a failure, as I have said, for the element of resistance against internal and external foes was weak. The French, who settled themselves in the bay of Rio de Janeiro, were expelled by the forces of the central government established in Bahia, and the city of Rio de Janeiro was founded in 1567 on a rock at the entrance of the bay, just as Bahia had been founded a few years before. So our first towns were fortresses established on the high ground for defensive purposes (Fig. 2).

The history of the seventeenth century in Brazil is not less connected with the topography of the country. European events had made Portugal a part of the Spanish empire between 1580 and 1640. The great struggle for the liberty of the seas, between the Dutch doctrine of mare liberum and the Spanish doctrine of mare clausum or monopoly, aroused the interest of the Dutch in the new Spanish colony of Brazil. Of course the economic development of the north drew their special attention, and in 1624 and 1630 the Dutch conquered Bahia and Pernambuco.

Before 1630 the colonial activity of Pernambuco, inhabited by seafaring peoples, had been a lateral expansion along the coast; thus Parahyba, Natal, and even Camocim were founded. After 1630 the economic necessity of conquering more inland pastures, of subduing the Cariris Indians, and especially of organising for defence against the Dutch, who dwelt on the coast, made the hinterland of Pernambuco a centre of resistance. After twenty-five years of struggle we were able to oust the Dutch from our shores. They had not come to South America with colonising ideas, but only for commercial purposes. They occupied a narrow strip of land on the coast (Fig. 1), and the topography of the Borborema highland allowed the Portuguese settlers, breeders, and soldiers to retire safely and prepare for victory against the invader. Professor Robert Rait wrote in his History of Scotland, "If any warfare

defensive purposes.

made Scotland a nation it was the struggle with Norway; and the victorious issue of that struggle in 1263 may be taken, from this point of view, as the completion and the seal of the consolidation of Scotland." The same words may be applied to Brazilian history. Our victory of 1654 against the Dutch gave us the consciousness of our nationality and sealed our political unity. Spanish America had no external foe to struggle with: that is why Bolivar's dream of unity never

came true.

The history of your western shores and isles was almost repeated on our eastern shores, because the same geographical conditions allowed us to organise the defeat of the invader.

For the sake of argument let us suppose that the Dutch episode of our history had been acted at the foot of the Serra do Mar, in the south. Two hypotheses are possible. In the first case, the Dutch would have given up their conquest at once, for the value of the coast would not have been worth heavy sacrifices, if the highland were to remain hostile. In the second case, if the Dutch had felt much stronger, they would have then organised the military conquest of the highlands of S. Paulo, where a very important Dutch colony would possibly still exist to-day. The enterprise would have been, to a certain extent, easy for them, for the southern settlements were still very weak, and Spain was not very keen in defending colonies that, after all, did not belong to her.

With the colonisation of the S. Paulo highlands by the diplomacy of the sugar planters of S. Vicente, and the colonisation of the Borborema highlands by the armed conquests of the cattle-breeders of the northeast, the historical part of the coastal mountains seems to end. But the importance of the Serras is still great and far-reaching in the economic history of the country, in the penetration of the interior by roads and railways, by foreign influences and activities, in the cost of freights, and in a thousand other details.

III. HISTORICAL PART PLAYED BY THE RIVERS.

At the beginning of the seventeenth century the penetration of the Brazilian interior was very restricted; from the Parahyba do Sul to the Rio Goyana, in Pernambuco, a strip of ten or twenty miles in width was the only occupied part of the country, and here sugar and cotton were grown.

We then begin to observe how, in the north, the needs of the cattlebreeding industry, and in the south the labour question leading to the enslavement of the Indians, and, later on, the lure of the mines, attracted populations towards the interior. In a new country rivers generally constitute lines of least resistance, and therefore are freely used by the colonists of the coast. Small countries with articulated coasts can easily forgo the advantage of big river systems, but large countries like Russia, China, India, the United States, or Brazil have their political and economic history connected with that of the great rivers.

"South America and Africa are alike in the unbroken contour of their coasts," writes E. C. Semple, "but strongly contrasted in the character of their rivers. Hence the two continents present the extremes of accessibility and inaccessibility. South America, most richly endowed of all the continents with navigable streams, receiving ocean vessels three thousand miles up the Amazon, as far as Iquitos in Perú, and smaller steamers up the Orinoco to the spurs of the Andes, was known to the explorers fifty years after its discovery. Africa, historically the oldest of continents, but cursed with a 'mesa' form, which converts nearly every river into a plunging torrent on its approach to the sea, kept its vast interior, till the last century, wrapped in utmost gloom."

On the other hand, looking at a map of navigable rivers in Brazil, we must admit that the striking feature is that, with the exception of the Amazon and the River Plate systems, there are no rivers representing a considerable extension of the seas, for as a rule the fall-line is not found very far inland. The S. Francisco river, for instance, offers *300 kilometres (about 185 miles), more or less, to navigation from its mouth to the first fall. The Parahyba do Sul has about 100 kilometres navigable in its lower part. Nearly all the Brazilian coastal centres are situated at the mouth of a river, but the fact is more often due to the