66 THE VARIATION OF ANIMALS IN NATURE 



which represents a stage in a particular lineage. Thus at the 

 stratigraphical level of the Millstone Grit, Carruthers found 

 two distinct species of coral, ^aphrentis constricta and £. disjuncta, 

 though each of these at this horizon represented a stage in an 

 individual lineage in which the individuals cannot be speci- 

 fically delimited from individuals that occur in earlier and 

 later horizons. It seems that the character-complexes, in 

 which the individual characters in any one lineage are modified 

 at different rates and so afford no regular correlation by 

 which species may be recognised, do in fact diverge so that 

 one lineage may differ from another at a given moment in the 

 same way as the species of the neontologist differ. In other 

 words, the investigations of lineages have revealed distinct 

 divergences equivalent to species, but these divergent groups 

 show no discontinuity in time from their predecessors or suc- 

 cessors. The criticism that the forms on which such studies 

 have been carried out are peculiarly plastic (Robson, 1928) 

 and therefore apt to be misleading has, we think, been suffi- 

 ciently answered by Trueman (I.e. p. 307), although there must 

 always exist some element of doubt as to the relationship 

 between groups diagnosed on certain plastic characters of the 

 shell and those founded on more stable characters. Finally, 

 it must be observed that the existence of lineages could be 

 suspected from the distribution of variants in modern popula- 

 tions (cf. p. 176, Chapter VI). 



III. Geographical Categories 



The subordinate units within the species recognised in 

 taxonomy and associated with the intensive study of geo- 

 graphical distribution are somewhat diverse and no standard 

 usage obtains. There are some outstanding works on the 

 geographical variation of single species or on allied forms, 

 such as those of Heincke (1898), Duncker (1896) and Schmidt 

 (191 8-1 930) (fishes) ; Sumner (1932) (Peromyscus) ; Crampton 

 (191 6-1 932) (Partula). Alpatov (1924, 1929), Semenov-Tian- 

 Shansky (1910), Rensch (1929) and others have attempted 

 to define the terms used. 



Mammalogists, ornithologists and, to some extent, 

 herpetologists regularly subdivide the species into subspecies 

 or smaller units such as races, all of which are characterised 



