CAUSES OF THE EVOLUTION AND EXTINCTION OF THE TITANOTHERES 



865 



had become thoroughly adapted; (2) it may intro- 

 duce new poisonous or deleterious plants; (3) it may 

 be the means of erecting new forest barriers or new 

 forest migration tracts for certain Carnivora; (4) it 

 may be the means of introducing new insect and other 

 pests, as well as new insect barriers; (5) it may be the 

 means of introducing new diseases and new insect 

 carriers of disease. 



Besides the changes in plant food that are brought 

 about by diminished moistui'e, as indicated below, 

 there are the effects of increased moisture which may 

 be equally if not more important. Moist conditions 

 may cause increase of certain forms of plant life dele- 

 terious to all Herbivora. Dry or moderately dry 

 conditions, provided they are not too extreme, are 

 generally more favorable to quadrupeds than moist 

 conditions. The plateau and forest regions most 

 densely populated with quadruped life, such as those 

 of Africa, are regions of moderate rainfall and even of 

 prolonged summer droughts. On the other hand, the 

 regions least densely populated with mammals are 

 those of heavy rainfall, dense forests and vegetation, 

 such as those of the equatorial belt of South America 

 and of Africa. 



It follows that periods of secular increasing moisture, 

 such as the warm interglacial stages of Pleistocene 

 time in the Northern Hemisphere, may have proved 

 unfavorable to certain large quadrupeds through in- 

 crease of humidity and rise in temperature even 

 prior to the advent of extreme cold. 



Insect 'barriers and moisture. — It is now a matter of 

 general observation that, especially in tropical, insect- 

 infested countries, dry seasons result in the reduction, 

 moist seasons in the increase of disease. Dry localities 

 are generally favorable, moist localities generally 

 unfavorable to quadruped life. For example, Shipley 

 (1906.1) observes of the tsetse fly that its 



northern limit corresponds with a line drawn from the Gambia 

 its southern limit is about on a level with the northern limit of 

 ■ Zululand. Most writers agree that the tsetse is not found in 

 the open veldt, that it must have cover. Warm, moist, steam}', 

 hollows, containing water and clothed with forest growth, are 

 the haunts chosen. 



This subject will be treated at length under insect 

 life as a factor in extinction. 



General ejfects of decreasing moistwe; secular desic- 

 cation. — We observe that deci'easing moisture (1) 

 changes the character of the food supply by diminish- 

 ing the softer and more succulent vegetation and in- 

 creasing the harder and more resistant vegetation; 

 (2) increases the length and severity of the dry 

 season; (3) removes forest barriers and admits new 

 competitors; (4) eliminates animals incapable of 

 traveling long distances for food and water, or living 

 on a limited or irregular water supply; (5) favors 

 grazing quadrupeds and eliminates browsing and 

 forest-living quadrupeds; (6) favors hypsodont (long- 



toothed) and is inimical to brachyodont (short- 

 toothed) quadrupeds; (7) favors cursorial (rapid- 

 moving) and is inimical to mediportal (slow-moving) 

 quadrupeds. 



Secular desiccation (Hann, 1903.1, p. 375) of certain 

 very extensive regions has occurred in parts of all the 

 great continents — North America, Australia, Asia, 

 and Africa — and on each continent we observe a 

 general concomitant modification and extinction of 

 certain kinds of quadrupeds. In general, there ap- 

 pears to have been a progressive decrease of moisture 

 in the Northern Hemisphere, beginning in Oligocene 

 time, which was checked only by the humidity of 

 preglacial time. Increasingly prolonged summer 

 droughts were characteristic of the late Miocene and 

 Pliocene of Europe, and we are beginning to ac- 

 cumulate evidence that the same conditions pre- 

 vailed in North America. 



The great regions of the world where decreasing 

 moisture has introduced a series of changes ending in 

 the extinction of a great number of quadrupeds are 

 (1) North America (Western Plains region, Interior 

 Basin, arid Plateau and Mountain region), beginning 

 in Oligocene time; (2) South America (Patagonia and 

 Pampean region), beginning in late Pliocene time; 

 (3) north-central Africa (the Fayum district), begin- 

 ning in Oligocene time; (4) central Australia, begin- 

 ning in Pleistocene time. 



American paleontologists, also Stirling (Australia), 

 Andrews (Fayum), and Ameghino (Patagonia), de- 

 scribe faunas adapted to much moister and more 

 hospitable conditions than those which now prevail in 

 the regions they consider. 



Sandstorms. — Bravard, who did much to explore 

 the Pleistocene fauna of South America, concurring 

 with Lyell, argued that the great South American 

 mammals were overwhelmed alive by moving sand, 

 as the simoom is said to overwhelm creatures in the 

 Sahara. He invoked for the purpose an exaggerated 

 form of the hurricane, still known in South America as 

 the pampero, and urged that, having been thus 

 killed, they were afterward covered by the sand. 

 Against this hypothesis Burmeister observes that the 

 greater number of the remains that are found are 

 isolated skeletons, most of which, curiously enough, are 

 without heads or tails. Such mutilations do not 

 sustain the idea that the creatures were overwhelmed 

 alive by moving sand. 



Secular desiccation and vegetation. — The indirect in- 

 fluences of dry secular changes of climate on quad- 

 rupeds are apparently quite as effective factors in 

 extinction as the direct influences, such as changes in 

 vegetation due to decrease of moisture, which make 

 certain types of quadrupeds that were perfectly 

 adapted to one kind of plant food largely or wholly 

 inadapted to the new or altered kinds of food. 



