224 ANNUAL REPOET SMITHSONIAN INSTITUTION, 1915. 



Let US imagine for a moment that we could entirely empty the 

 Atlantic Ocean, drain it completely dry ; and, that done, let us con- 

 template from above the relief of its bed. We shall see two great 

 depressions, two enormous valleys extending north and south, par- 

 allel with the two shores, separated one from the other by a median 

 zone elevated above them. The western valley, extending the length 

 of the American coast, is the larger and deeper of the two ; it shows 

 oval fosses, or depressions, descending to more than H,000 meters 

 below the level of the shores, and also occasional elevations, one cor- 

 responding to the Bermudas, which, from the bottom of the gulfs, 

 rise boldly toward the surface. The eastern valley, along the Euro- 

 pean coast, appears to us narrower and of less depth, but much morp 

 hilly; and numerous pyramids, some thin and slight like that of 

 Madeira, others massive like those which bear the archipelago of 

 the Canaries and Cape Verde, rise here and there in the midst of 

 the valley or near its eastern border. The much elevated median 

 zone outlines before us a long promontory, whose axis coincides with 

 that of the Atlantic abyss. It curves in an S shape like the two 

 valleys and the two shores, and, starting from Greenland and sur- 

 rounding in its mass Iceland and the northern islands, goes tapering 

 southward and ends in a point at the seventieth parallel of latitude 

 south. In most of its course, this promontory has a mean breadth 

 of about 1,500 kilometers (937.5 miles). Far from being regular 

 and with a uniformly spherical curve, its surface is everyAvhere 

 indented, bristling with projections, riddled with hollows, especially 

 in the region of the Azores, what we call the Azores being merely 

 the summits of the highest protuberances. 



In this complete view of the ocean drained and dry we would 

 certainly observe many other things, which are otherwise invisible 

 beneath the w^aters. We would see not only the longitudinal ar- 

 rangement which I have just described and which has been revealed 

 to us by soundings but also the transverse irregularities w^hich can 

 not fail to exist and of which, at the present time, we know 

 almost nothing, because the soundings have not yet been numerous 

 enough. The map of the archipelago of the Azores shows clearly 

 that the nine great islands of which it is composed are ranged along 

 three parallel bands, in a direction from east-southeast to west- 

 northwest; and these bands are staked out by the islands over a 

 total length of nearly 800 kilometers (500 miles). No doubt such 

 lines are prolonged very far under the waves, and they would have 

 great importance in making a model of the ocean bed, but they are 

 evidently not the only ones. The day will come w^hen the charts of 

 the Atlantic depths wall be exact and detailed; we shall then see 

 fault lines and bands of folds crossing the vast abyss and extending 



