4 J. C. BRANNER — STONE REEFS ON COAST OF BRAZIL 



with it. Usually they have one end connected with the land, at least 

 during low tide, while the free end stands across the mouth of stream, 

 embayment, or estuary; sometimes, however, they are not connected 

 with the shore at all, at any stage of the tide. At high tide they are 

 about flush with the surface of the water, while at low tide they are 

 exposed like long, low, flat- topped walls or breakwaters. In width they 

 vary from a few paces to 450 feet ; in length they are from a few hun- 

 dred feet to 8? miles, the total length being concealed by sand, which 

 covers one end. Commonly they are almost straight ; when they curve 

 at all the curves are gentle and only perceptible when one looks along 

 them lengthwise or when they are carefully mapped. 



These remarkable natural walls of sandstone accompany the shores of 

 northeastern Brazil, with many interruptions, from Ceara to Porto Se- 

 guro, a distance of 1,250 miles. With unimportant exceptions, they do 

 not occur beyond these limits. 



Factors in the Formation of the Reefs 



in general 



Having stated the broad theor}^ of the origin of these stone reefs, I shall 

 not weary the reader with the tedious process of elimination by which 

 the accepted conclusions were reached, but shall invite attention direct^ 

 to their history as finally worked out in a study of the problem which 

 has been carried on with various interruptions for nearly thirty years. 



HISTORY OF THE COAST 



It will be necessary first to briefly outline the geologic and geographic 

 history of this particular coast. The greater part of the rocks of the 

 coast region are marine sedimentary beds, apparently of Eocene age- 

 During Miocene times the coast seems to have stood several hundred 

 feet higher than it does at present, and narrow valleys were cut in the 

 Eocene beds, mostly at right angles to the coastline. This period of 

 erosion was followed by a depression, when many new bays were formed 

 by the valleys which lie near the coast. There have been some changes 

 of level since this period, but they have not been great, the evidence 

 nowhere suggesting a movement exceeding 30 or 40 feet. In the mean- 

 time the strong on-shore waves and the in-shore currents rapidly cut 

 away the soft Eocene sediments of the headlands and threw them into 

 the reentrant angles of the coast. This process continued until the 

 mouths of the bays were completely or nearly closed and the entire coast 

 made as nearly straight as it is possible for as long a coast to be. Streams 

 flowing into these narrow bays silted them up from the upper ends. 

 Marine erosion was so vigorous, however, that several such bays were 

 almost completely closed long before the land sediments filled them up- 



