260 I The Process of Evolution 



reptiles into terrestrial and aerial lines (represented now by the 

 crocodiles and birds) to the formation of species of mammals. On 

 the other hand, some lineages apparently did not split for long 

 periods of time but underwent such gradual morphological change 

 that taxonomic distinction anywhere along the continuum is virtually 

 impossible, except at the beginning and the end. This evolutionary 

 mode is called phyletic evolution. Simpson recognized also a third 

 mode, quantum evolution, which is rapid change involving the ac- 

 quisition of characteristics for a completely different way of life. 

 The partial or complete joining of phylogenetic lineages is difficult 

 to recognize in the fossil record, although presumably it played the 

 important role in the past that it does today. 



One of the most interesting problems that can be studied with the 

 fossil record is the rate at which evolution has occurred. There are 

 many diflFerent ways of measuring rate of evolution. At present none 

 of them is entirely satisfactory, but nonetheless they are of great 

 interest. In many of the well-worked-out lineages of vertebrates and 

 mollusks, rate of change in particular characters can be studied with 

 time as a dimension. It is also possible to measure the rate of "forma- 

 tion" of taxonomic categories, assuming that these are proportional 

 to some stage of morphological distinctness or some degree of genetic 

 isolation. The rate of extinction of taxonomic groupings, as expressed 

 in survivorship curves, also may be of use. It must be remembered, 

 of course, that comparisons based upon taxonomic distinctions be- 

 tween groups must be interpreted with great caution. There is ample 

 evidence from the study of living and fossil forms that what is called 

 a genus, say, in one group may not be equivalent biologically to 

 what is so designated in another. 



The multivariate-correlation analyses now being employed in nu- 

 merical taxonomy ( Chap. 13 ) seem to offer real possibilities for the 

 development of more satisfactory methods of measuring evolutionary 

 rates. By studying changes through time in a large sample of char- 

 acters, objective estimates of the total amount of structural diver- 

 gence may be obtained. The smaller the degree of correlation be- 

 tween successive samples, the greater the amount of evolution. In 

 such an analysis, arbitrary taxonomic judgments can be avoided. 



The measurement of time also is beset with difficulties. While 

 there are modern methods of dating which presumably give reliable 

 results, there is not general agreement concerning the beginning or 

 the duration of the major geologic periods. Age determinations have 

 the same sampling problems as studies of fossils and must be corre- 

 lated with the occurrence of known fossils. Experimental error may 

 be relatively large. In general, establishment of an absolute time 



