32 PR0CEEDIXG3 OF THE 



appear to diHer from lliose of tlie hitr-r llyiiif;; reptiles in being 

 relativ el}' larg*^. As the group is Iraceil iipwiirils in time some of 

 its nu'nil)ers are progressively larger, until just before its extinc- 

 tion at the end of the Cretaceous period it is represented by 

 I'tcranodoii, with a \\ing-span of '20 feet or more, the largest 

 Hying animal wiiidi ever existed. In the course of tbis evolution 

 S'jme of the I'terosauria become toothless, like Pteranodon itself, 

 but the only essential change is a firmer tixatio)i of the wings 

 and a reduction of the clawed digits to mere splints on the wing 

 finger. From its very b( ginning, so far as our present knowledge 

 extends, the Pttrosanrian was an eliicient glider or flier, and no 

 real due to its ancestry has hitlierto been fonnd. 



Finally, consider the origin of tbo Maiuinalia. Palieontologists 

 pride tliemselves on baving found during recent years almost 

 every gradation between certain skelt-tons of the Permian and 

 Triassic Tlieroniorph reptiles of Soutb Africa and the skeleton of 

 the monotrenie mammals. The series is, indeed, now remarkably 

 com|)lete and convincing. There can be no doubt, from the dis- 

 covery of a few isolated jaws and limb bones both in Europe and 

 North America, that representatives both of these lowly mammals 

 and even of n)arsupials were widely spread during the latter bait" 

 of the Mesozoic era. They are still better known by compara- 

 tively specialised forms in deposits dating back to the very 

 beginning of the Tertiary era both in Europe and in North 

 Auierica. Nearly all of them then sudilenly disappear in both 

 countries, and they are replaced by typical placental mammals, 

 already differentiated into several modern groups, wbiih seeni to 

 have no connection with the primitive n;ammals which immediately 

 ])receded them in the same regions. At tbis stage, therefore, 

 there is a complete gap in the series ; and even if the earliest 

 mannnals can be clearly traced back to antecedent reptiles in 

 South Africa, the bigl)er mammals characteristic of the present 

 world still appear without any recognisable ancestors. 



A possil)le explanation is that each new lineage began as a 

 rapid development in one community in a locality of restricted 

 extent. In this case its initial stages would be represented by 

 comparatively few fossils in a small area, or none of them may 

 have been preserved owing to local unfavourable circumstances. 



It is already known that there were such rapid local develop- 

 ments. The earliest stages of the elephants, for example, are 

 found only in Egypt, and they are represented by so variable a 

 series of forms that it is dillk-ult to classify them. They are also 

 accompanied by strange mammals more or less related to ancestral 

 elephants. The region of which modern Egypt is now a part 

 seems, therefore, to have been the place of origin of the ele])hant- 

 lineage. Another strange and rapid development occurs in the 

 ^fiddle Eocene rocks of Wyoming, where the small-brained Diiio- 

 cerata suddenly arose in many forms, but as suddenly died out 

 without ever spreading. The local development and sudden end 



