2i4 SCIENCE PROGRESS 



Such is the primary aim of palaeontology, the tracing of 

 lines of descent. The same object lies before the taxonomist, 

 but difference of material leads to a corresponding diversity 

 in method. The two students differ, as do the historian who 

 has as his material documents the age of which is known and 

 the modern social anthropologist who must reconstruct the 

 past from the actually existing organisation, customs, and 

 traditions of his people. 



The historian and the palaeontologist by their direct and 

 contemporary evidence have certainty of order so far as their 

 material extends, and an abundance of detail ; the anthro- 

 pologist and the taxonomist can only reconstruct the broad 

 outline of history and are at best uncertain of the relative 

 periods at which changes in structure, of society, or of an 

 organism respectively have come about. 



Were the preserved contemporary documents complete, 

 covering the whole field of study, the construction of a history 

 would be a simple process ; but their fragmentary nature, with 

 whole periods unrepresented by documents, and many even 

 of those which have been spared to us mutilated, renders their 

 direct interpretation difficult. 



The incompleteness of individual items cannot be over- 

 come — we have to take them as we find them ; but for the 

 filling in of the gaps we can fall back on the methods of the 

 anthropologist or the taxonomist. 



These methods depend on the fact that indelibly im- 

 pressed in structure are the traces of former conditions. The 

 full realisation of the truth of this doctrine, so far as it concerns 

 biology, was due mainly to the work of two palaeontologists, 

 A. Hyatt and L. Dollo, to the latter of whom we owe its clear 

 statement and a demonstration of the possible extent of 

 its use. 



The method of Alpheus Hyatt depends on an application 

 to fossil material of the theory that " ontogeny repeats phy- 

 logeny," the view perhaps first stated by Louis Agassiz, that 

 an animal in its youth repeats the adult conditions of its 

 ancestors. When applied, as it always has been by embry- 

 ologists, to the reconstruction of structures occurring in very 

 remote ancestors, this theory has seldom proved useful (its 

 most striking success being Reichert's theory of the mamma- 

 lian auditory ossicles), because early stages are usually very 



