71 



lands, which would then form islands, on the borders of which they 

 would congregate in multitudes, until numbers of them, collected in 

 ravines, rushing on in their attempts to escape in search of food, 

 might trample each other to death and leave their bodies at the foot 

 of impassable precipices washed by this new sea. 



Conjecture II. Let us suppose that the basin of the Mediter- 

 ranean was formerly dry land shut in from the Atlantic by a barrier 

 at the Straits of Gibraltar. In these circumstances the whole of the 

 present Mediterranean Sea would have been a country sunk more or 

 less under the level of the ocean ; just as a large tract of country 

 around the Dead Sea is at present. 



The country included in this great depression might be warmer 

 than others situated in the same latitude, and would be full of lakes 

 fed by rivers, since there could be no drainage except from evaporation. 

 It would therefore probably be favourable to the growth of the 

 Hippopotamus. 



Imagine now some convulsion to have opened the Straits of 

 Gibraltar, so as to have allowed the waters of the Atlantic to enter 

 this vast basin. The salt river thus introduced might require days, 

 or weeks, or months, or years, before it filled this immense cavity ; it 

 might also increase its velocity as it wore away the channel of its 

 entrance. 



Under these circumstances the Hippopotami would be gradually 

 driven to the higher ground, and those mountainous regions which 

 now form the islands of the Mediterranean would then have re- 

 ceived on their shores the hosts of animals, driven by this inundation 

 to seek dry land on which to repose and food on which they might 

 subsist. 



Both of these hypotheses account for the aggregation in different 

 localities, of the remains of large numbers of animals of the same 

 class, dwelling amongst rivers and lakes ; both equally account for 

 their destruction on the same spots, either by trampling each other to 

 death in the rush to escape, or by the slower processes of drown- 

 ing or starvation. But neither hypothesis accounts for the fractured 

 state of these bones, even though the animals should have rushed 

 over the precipice. It might be expected that some portion of these 

 Hippopotami, escaping from the deluge which destroyed their race, 

 would have reached the plains of some larger island, and then 



