146 T O R R E Y A 



have unusually broad tolerances. Circumstantial evidence, on the contrary, in- 

 dicates that old relic species are frequently markedly restricted in area and 

 habitat type. 



This problem has received at least one excellent consideration in paleo- 

 botanical literature. Ater pointing out that certain fossil floras of later Tertiary 

 age contain mixtures of plants from widely different habitats, Axelrod (1941) 

 suggests that the explanation may not be due only to overlap of floras (in 

 ecotonal regions, or from migratory mingling), or to the fact that Miocene 

 and Pliocene vegetation was "generalized" and modern forests derived by 

 "climatic segregation only in the late Cenozoic," but to the ancient existence 

 of ecospecies. For example, Sequoia Langsdorfii (close to ^. sempervirens) 

 was variously associated with species of boreal, warm-temperate, and temperate 

 type. Other modern endemics, now of restricted type, but of once wider asso- 

 ciation, include Lyonothammis, Ginkgo, Glytostrobus pensilis, Picea Brew- 

 eriana, and Quercus tonientosa, according to Axelrod. He says, "it seems 

 highly probable that many Miocene and Pliocene species related to living en- 

 demics may represent extinct ecotypes of more widely distributed Tertiary 

 ecospecies." Probable as this concept is, it still does not show that primitive 

 species are of wide ecological tolerance and recent ones of narrow amplitude. 

 The late Cenozoic was a time of climatic breakup and, for many species, bio- 

 type depauperization with only "senile," relic endemics remaining; but, as 

 Axelrod supposes, the wide area and diversified conditions under which cer- 

 tain Tertiary species lived were due to the biotype (ecotype) richness of the 

 species as a whole. That richness represents the mature condition of a species 

 history. As with previous criteria, we find ourselves confronted by many "ifs." 

 The above arguments concerning the region of least dependence upon a re- 

 stricted habitat are applicable in the determination of center of origin only 

 when the center of origin is also the center of variability, and when the center 

 of origin has not been disturbed and reduced in biotype richness. 



The idea that a species is usually ubiquitous in the center of its range, oc- 

 curring in all kinds of places, and restricted to only the most favorable sites 

 at its areal limits, according to Griggs (1914), is probably attributable to 

 Blytt, and has been favored by Cowles. This idea includes the assumption that 

 the favorable climate in the central portion of the species range somehow over- 

 comes diverse edaphic factors, whereas at the margin of range edaphic factors 

 permit a spotty extension of area. I remember Cowles, when lecturing on the 

 dunes of Lake Michigan's shores, saying of the cactus, "It sits on the south- 

 ern and western slopes, looking toward its home." There is, of course, a large 

 element of truth in this generalization, as is shown by the usual disposition 

 of preclimax and post-climax communities in any region. 



