216 SCIENCE PROGRESS 



of those distinct but nearly allied phyla which are the smallest 

 groups we need recognise. 



The phylogenies so constructed, broken and incomplete 

 as they must always be, and inaccurate either in grand features 

 or in detail as they are, make closer and closer approxima- 

 tions to the truth as reliable material becomes more abundant 

 and as the methods of its study become more refined. 



Such histories, whose formulation is the primary aim of 

 the palaeontologist, may be used in many ways. They serve 

 as a check to evolutionary theories, bringing them to the bar 

 and confronting them with new series of the actual facts 

 which they are intended to explain. They enable us to see 

 and study the gradual improvement of an animal mechanism ; 

 to trace how, whilst retaining its capacity for performing its 

 function at every stage, it gradually changes until it may be 

 adapted to some quite new and highly special purpose. 



The many phylogenies which have been studied by palaeon- 

 tologists during the last thirty years, whether they were 

 actually expressed or remained implicit in the published 

 work, have many features in common, features which distin- 

 guish them from the early efforts of zoologists and which 

 remained unsuspected until they were abundantly revealed by 

 palaeontological evidence. 



Palaeontological phylogenies are of the most diverse kinds : 

 they may represent the real development of some very re- 

 stricted group, or they may deal in a broad way with the 

 development of orders and superorders. They have been 

 investigated in all groups of animals both vertebrate and 

 invertebrate. 



The fact that students of echinoderms and vertebrates 

 produce phylogenies of similar type, and find that these agree 

 in character with those which represent the histories of groups 

 of cephalopods and brachiopods, shows that the underlying 

 factors which have determined the evolution are similar in 

 all groups. All modern carefully investigated phylogenies 

 founded on palaeontological data agree in the following broad 

 features : 



i. That in any one line evolutionary change, especially of 

 those regions of the body which do not seem to show adapta- 

 tion to any special mode of life, proceeds steadily, no matter 

 what the stock's changes of habit, as if from the first devoted 



