77o CONCLUSION xxxn. 4- 



study of the existing forms, because the birds are conspicuous, well 

 known, and varied (p. 522). 



A proper appraisal of the nature of evolutionary change demands an 

 understanding of the fact that evidence about it comes from very 

 different sources and varies in different animal groups. We may there- 

 fore profitably extract from the results of various parts of our study 

 such simple propositions as are strictly justified by the evidence and 

 provide us with a sure foundation of knowledge about the subject. 



5. Rate of evolutionary change 



It has recently become possible to consider several ways of measur- 

 ing the rate of evolutionary change. Simpson (1953) distinguishes 

 between measurement of (1) genetic rates, (2) morphological rates, 

 (3) taxonomic rates, and we may add as a possibility (4) rates of change 

 of information flow. Although the first and last of these are biologically 

 the most instructive, they are impossible to measure on a large scale or 

 in extinct populations. Rates of change of linear or other dimensions 

 can be estimated in suitable cases such as horses' teeth (p. 738). 

 Haldane suggests that changes should be considered on a percentage 

 rather than an absolute basis, for instance by considering the time 

 needed for a unit increase in the natural logarithm of a variate or one 

 standard deviation. Change of one s.d. per million years might be 

 called a 'darwin'; the horse tooth change (p. 737) being then at a 

 rate of 40 millidarwins. 



The only easily available quantitative data about large groups of 

 animals are the number of taxonomic units (species, genera, &c.) into 

 which they are divided. If it were true that differences between, say, 

 genera, meant the same when used by different authors and in different 

 animal groups then we could measure rates of evolutionary change by 

 the numbers appearing at each taxonomic level. The condition is 

 unfortunately not strictly fulfilled and there are inevitably examples of 

 what has been called 'monographic evolution'. Nevertheless, the defini- 

 tion of a difference as of, say, 'generic' or 'ordinal' rank by a competent 

 worker is in effect a kind of measure of general morphological difference. 

 Indeed in view of the subtle efficiency of the human receptors and 

 brain for this sort of comparison, the measure is perhaps as accurate 

 as could be expected to result from any artificial 'morphometer' 

 that can at present be imagined. 



Using taxonomic criteria it is clear that rates of evolution vary in 

 different groups. Thus the living prosimians have changed little since 

 the Eocene while their descendants have gone on to produce the whole 



