J. IMBRIE 



133 



sifications of living and extinct species are on an entirely differ- 

 ent level. For the vast majority of genera commonly found as 

 fossils, however, this judgment is exaggerated or untrue. There 

 can be, of course, no conflict between the classification of living 

 and extinct Euglena: like hosts of other soft-bodied forms this 



-.23-.22-.2l -.20 -.!■» -.18 -.17 -.16 -.15 .14 -.13 -.12 -.11 -.I0-.C -J38-.0? -.0* -.03-.0* -.03 -.02 -.01 .01 .02 .03 £4 .05 .06 .07 .08 M .10 



Fig. 4. Ratio diagram showing comparisons of certain osteological char- 

 acters in two species of the modern marten Mustek, and two modern sub- 

 species of Mustela sibirica. The latter show a strong overlap of measured 

 characters, whereas the two species show little overlap in the same charac- 

 ters. A, Mustela sibirica fontanierii; B, M. s. davidiana; C, M. altaica ka- 

 thiah. Abcissal scale represents for each character the difference between 

 the logarithm of any given value and the logarithm of the mean value of 

 the population selected as reference standard (in this case, population A). 

 Thus the horizontal distance between any two points represents the ratio 

 of either one to the other. Three points connected by a horizontal line are 

 plotted for a given character in each sample to represent the greatest, least, 

 and mean dimensions. The several means for each character in a popula- 

 tion are connected by dashed or solid lines. For discussion of ratio dia- 

 grams, see Simpson (1941). Mammal data from Allen (1938) (Colbert 

 and Hooijer, 1953). 



genus leaves essentially no fossil record. Furthermore, it is gener- 

 ally true that specific discrimination in groups of animals com- 

 monly found as fossils is ( or can be ) based on skeletal morphol- 

 ogy. Consider, for example, the shell-bearing foraminifera, radio- 

 larians, corals, ectoproct bryozoans, brachiopods, snails, clams, 



