558 PALEONTOLOGY 



Ceratodus was also universally distributed in the waters, probably 

 even in the seas, of the middle part of the Mesozoic epoch. So that 

 in each of these three cases Australia and South America may be 

 merely refuges for old forms of life which were lost much earlier by 

 extinction in other parts of the world. They need not have been 

 directly connected. 



In short, when using land animals or fresh-water animals as tests 

 of former changes in the distribution and connection of land areas, 

 it is necessary to make a distinction between those of restricted range 

 and those of past or present cosmopolitan distribution, the former 

 alone affording reliable evidence. 



Relations to Biology 



It is already clear that the scientific value of a fossil depends upon 

 the exactness with which the circumstances of its discovery are 

 determined by a geologist. The briefest experience is also enough'to 

 demonstrate that the well-mineralized remains of an organism can 

 only be satisfactorily interpreted by an observer who is familiar 

 with the structure of rocks and their common constituents. The 

 student of fossils needs as much elementary training in the geological 

 succession of the rocks and the varied nature of mineralization as 

 the student of histology and embryology requires to locate his 

 sections with exactitude, and to understand the action of the different 

 stains and media he employs. In the one case nature makes the 

 preparation; in the other case the processes of laboratory technique 

 are responsible for the difficulties. In both cases, there is scope for 

 numerous fantastic conclusions if the properties of the preservative 

 medium are misunderstood. 



Paleontology, however, is essentially a department of biology, 

 and it can only be prosecuted with success by a skilled biologist, 

 who has had the elementary geological and mineralogical experience 

 just mentioned. It bears, indeed, the same relation to the whole 

 world of life that embryology bears to the structure of an individual 

 organism. The one deals with the rise and growth of races and their 

 varying relationships; the other describes and interprets the evolu- 

 tion of an individual and the processes by which the different parts 

 of its mechanism are finally adjusted. Both unfortunately depend 

 on extremely imperfect material ; for fossils are nearly always mere 

 badly preserved skeletons, and they represent only an infinitesimal 

 fraction of the life that has passed away, while embryos are so much 

 adapted to the peculiar circumstances of their environment that 

 many of the essential stages in their growth and development are 

 obscured and modified by temporary expedients. 



The past history of the world of life, as revealed by fossils, has 



