CH. XX] THE ORIGIN OF SPECIES 221 



and the parent will probably survive in the outer portions of its 

 range at any rate. 



It is clear, if the Age and Area explanation of the facts 

 of distribution be accepted— and as yet no other satisfactory 

 hypothesis is forthcoming— that the endemics must in general 

 be younger than the "wides," and it seems natural to suppose 

 that they have been derived from the latter. But if this be so, 

 then both parent and child occur together, or near together, in 

 most cases, and if one push this consideration to its logical con- 

 clusion, one will see that there is no reason why the whole tree 

 of the evolution of a genus (or even family) should not survive 

 upon the earth at the present moment, as I have contended for 

 the last fifteen years (120). Destruction such as that wrought 

 by the glacial periods, or other geological convulsions, might of 

 course kill out genera or families, but so long as conditions 

 remain reasonably constant, there seems no reason why they, 

 or intermediates, should be killed out. 



If, as seems probable, destruction in the struggle for existence 

 is to fall largely out of consideration as potent in the evolution 

 that has gone on (except that it must have destroyed tens of 

 thousands of incipient species, many of which might have been 

 of great value had man been there to preserve and investigate 

 them), we cannot regard Jordanian species as stages in the evo- 

 lution of Linnean, for to get the localised Linnean from Jordanian 

 species, Avholesale destruction must have gone on, killing out 

 altogether many of the latter. 



\Vhilst the exclusion of advantage to the species as a serious 

 factor in its evolution (though of great importance in deter- 

 mining whether or not it shall survive) practically compels us 

 to accept the theory of mutation, and that such as may give rise 

 at once to Linnean species, it also seems to me, when taken in 

 conjunction with other phenomena which are now clearly visible, 

 to involve other changes in our views. Chiefly important among 

 these is the new view of evolution, first proposed by Guppy in 

 1906, and by the writer in the following year, that evolution did 

 not proceed from indi^^idual to variety, from variety to species, 

 from species to genus, and from genus to family, but inversely, 

 the great families and genera appearing at a very early period, 

 and subsequently breaking up into other genera and species. The 

 final results of the study of Age and Area, with its demonstration 

 of the universality of the hollow curve, seem to me at present 

 almost to involve the acceptance of this view, and the subject 

 will be fully developed in a subsequent book. 



