38 GENERAL PRINCIPLES OF ZOO LOG Y. 



science, we invoke for the explanation of natural facts only 

 the forces of nature, we are driven to the hypothesis of 

 spontaneous generation, namely, that from materials with- 

 out life, by a peculiar combination of them, the compli- 

 cated mechanism which we call "life" has arisen. This 

 hypothesis also supposes that the first organisms possessed 

 the simplest conceivable structure. 



Variability not proven to be a Universal Principle. 

 Starting from a basis of facts, by generalization of the re- 

 sults we reach a simple conception of the origin of the 

 animal kingdom, but we have in equal measure departed 

 from the results of direct observation. Observations only 

 .show us that species are capable of changes and can from 

 themselves produce new species. That this capacity for 

 variation is a universal principle, a principle which explains 

 to us the origin of the animal world, needs further demon- 

 stration. 



Proofs of Phylogeny. The rise of the now living ani- 

 mal world is a process which has taken place in the thousands 

 of years long past, but is no longer accessible for direct 

 observation, and therefore it can never be proved in the 

 sense that we explain the individual development of an 

 organism. In regard to the conception of the simple evo- 

 lution of animals we can attain merely to proving the 

 probability ; yet it is shown that all our observations of 

 accessible facts not only agree with this conception, but 

 find through this their only simple explanation. Such 

 facts are furnished to us by the Systematic Arrangement 

 of Animals, Paleontology, Animal Distribution, Compara- 

 tive Anatomy, and Comparative Embryology. 



(/) Proofs from Systemization. For a long Mme it 

 has been recognized, and in recent times finds ever-increasing 

 confirmation, that if we wish to express graphically the 

 conditions of relationship of animals, their classes, orders, 

 genera, and species, simple co-ordination and subordina- 

 tion are not sufficient, but one must select a treelike ar- 

 rangement, in which the principal divisions, more closely 

 or distantly related to one another, the branches, phyla, or 



