36 KADIOGENESIS IN EVOLUTION. 



day species. We know the lineage of the horse, the elephant, 

 the camel and several other groups. Although there are 

 lacumie in our palaeontological knowledge we can never 

 hojie to fill, some strata have yielded surpri.singly complete 

 evidence on the evolution of certain groups. 



Still more familiar examples to Australasian Morker.'- 

 of radial evolution may be noted. In New Zealand ^^■ingless 

 avian forms developed on such lines as though they had 

 striven to take the place of the missing laud mammals. These 

 existed until recent times, and their pigmy cousins (the 

 Kiwis) are with us still. In Australia among the living 

 and extinct marsupial forms, the majority of the types 

 found in the true placentals have their analogues. We 

 have, or have had, large and small herbivorous, insectivorous 

 and carnivorous foriiis, and there are diverse examples of 

 serai-subterranean, burrowing, arboreal, parachuting and 

 saltatorial habits. Divergencies may be illustrated by the 

 dental characteristics, say, of Thylacoleo, Sarcophilus, 

 Hypsiprymnus, Diprotodon, Phascolomys. Myrmecobius and 

 Notoryctes. Such striking distinctions in an order in one 

 region may be generalised as due to a principle of Radio- 

 genesis. These marsupials have not been stationary since 

 Tertiary times ; they show processes of rapid evolution^ 

 and in some cases we have evidence of degeneracy ; nor 

 can we translate the characteristics of many of our extinct 

 forms in terms of lineal relationship with the life of to-day. 

 and they add further complexity to our fauna. 



Thus the hfe of the past makes multifold the variation 

 of to-day. Perhaps the crux of the whole problem may 

 be expres.sed in the two questions put by Osborn. ' the 

 American palaeontologist.* ^' Is it true that the greater 

 number of new or germinal characters which appear are 

 orderly and according to some entirely unknown law of adap- 

 tation ? Or is it true that the greater number of new 

 characters are accidental, disorderly, fortuitous, adaptive 

 or inadaptive. fitted or unfitted, and that order comes out 

 of chaos by the selection of those which happen to be fit I '^ 

 —natural selection mimicking design, as Balfour puts it. 



*Jour. Acad. Nat. Sci., Phil., Ser. 2, Vol. XV., 1912, p. 301. 



