THE ORIGIN OF SPECIES 



islands. Thus a sweepstakes bridge was formed, a bridge likely to be 

 crossed only by animals which could be carried from island to island by 

 such accidental means as floating on driftwood. Naturally, only a rela- 

 tively few small mammals succeeded in crossing. Rodents of the guinea pig 

 type (cavioids) and monkeys were the principal "island hoppers." But the 

 islands were again submerged in the late Oligocene. During the Pliocene, 

 the elevation of the land again established a chain of islands, but this 

 time the elevation proceeded until a continuous land connection, which 

 has persisted to the present, was established in the Pleistocene. Over this 

 bridge, a wide variety of mammals have passed. From North America to 

 South America went such mammals as deer, camels, tapirs, horses, masto- 

 donts, cats, weasels, raccoons, bears, dogs, mice, squirrels, rabbits, and 

 shrews. But other North American mammals were "filtered out," because 

 of ecological or topographical barriers which generally prevented their 

 gaining access to the bridge. These included such animals as pocket 

 gophers, beavers, bobcats, bison, and sheep. But porcupines, armadillos, 

 capybaras, and ground sloths, as well as other mammals, invaded North 

 America from the southern continent. Representatives of the first two 

 types have survived in North America, but the others have long since 

 become extinct. Simpson has said that there were twenty-nine families of 

 land mammals in South America and twenty-seven in North America 

 before the Pleistocene connection was established, with only two of these 

 families common to the two continents. This makes a total of fifty-six 

 families in the Americas. Soon after the connection was established, there 

 were no less than twenty-two families in common. Fourteen of these were 

 originally North American, and seven were originally Soiith American, 

 while one was of uncertain origin. Nine families were still confined to 

 North America, while seventeen were still confined to South America. 

 Thus the total was only forty-eight families, as eight had already become 

 extinct. At present, there are fourteen families in common, nine confined 

 to North America, and fifteen confined to South America, for a total of 

 only thirty-eight families in the Americas. Thus extinction has continued 

 since the Pleistocene. This filter bridge, then, has resulted in an extensive 

 exchange between the two faunas, but it has not in any way merged them 

 into a single fauna. This example has been discussed in terms of mam- 

 mals, because the mammalian paleontology for the Americas has been 

 intensively worked out, but it is probable that comparable results would 

 be obtaincxl for any group if adequate data were studied. 



Sweepstakes Routes. Sweepstakes routes are much less tangible be- 

 cause no actual land connection is present. It has often been stated that 

 natural rafts— driftwood or uprooted trees, for example— might carry 

 plants, particularly as seeds, and animals from one place to another. This 

 is the type of transportation most commonly envisaged in connection with 

 sweepstakes routes. But the concept need not be restricted to water bar- 

 riers. It can apply to any type of barrier if the crossing is improbable but 

 not impossible. Simpson has summarized the characteristics of a sweep- 

 stakes route as follows: Generally, only small animals, and particularly 

 arboreal types, can cross. The chances are much greater for some of these 



S18 



