OLIVER OX THE ATLANTIS HTPOTUESIS. 151 



objected that the fossil is ofteu referred first to the species, then to 

 the genus, and not as in recent botany, first to the genus, then to the 

 species ; but recollecting how seldom fossil remains enable us to 

 ascertain how far tr^'o forms may be removed in floral or in fruit 

 structure, which resemble each other precisely in their leaves ; see- 

 ing, moreover, that if the specific determination or analogy be correct, 

 that of the genus must necessarily be so, while indeed, if the former 

 be incorrect, the latter may yet hold good, — I cannot but think it 

 the safer coiu-se in the discussion of the present question to confine 

 myself to comparisons of genera solely. The case is Avidely difl:erent 

 when we compare the species of recent floras with each other, inas- 

 much as here we deal usually with indi\ddual elements of a value 

 much more nearly equal, and are thus in a position better able to 

 appreciate the minor facts of recent migration and modification of 

 type which such a collation might indicate, than it is possible we 

 could be from the comparison of a recent with a fossil flora, or of 

 two recent floras based ujdou their genera solely. In spite of their im- 

 perfection there can be little question, but th&t the most important 

 problems of plant-distribution are to be solved only by a constant 

 reference to fossil remains, and according as we compare existing 

 with extinct floras of recent or of more remote geological date, shall 

 we find that the comj)arison of species with species, of genus with 

 genus, of order with order, and of sub-kingdom with sub-kingdom, 

 have each their proper place and value in helping us to a right 

 ajDprehension of the changes which in respect of plant-distribution 

 our planet has experienced. In the case of the flora of the tertiary 

 period, from the imperfect nature of the evidence upon which we 

 must at present depend, and the cuTum stance that probably at least 

 one-fifth or one-fourth of its generic types, referable with more or 

 less probability to existing natural orders, are extinct or indeter- 

 minable, (exclusive of the various forms grouped under Fhyllites, 

 Antholites and Carpolithes) , I believe that a comparison of specific 

 forms is quite as calculated to mislead as reliably to inform ; and 

 although I regard Professor Heer's attempt to indicate the living 

 analogues of Swiss tertiary plants in his tabulated enumeration as 

 very able, yet I do not think the general results attained by it 

 add to the issue of a generic correlation ; while Professor Unger's 

 catalogue of tertiary species and theii' North American representa- 

 tives* appears to me overstrained in favour of the Atlantis hypothesis, 

 and calculated to give a false impression. We must not, however, 

 overlook the peculiar and qualifying circumstances, referred to above, 

 under which the generic determination of not a few fossil species 

 must be made : — that the reference of the fossil to a recent genus 

 frequently depends less upon the recognition in the fossil, of any one 

 essential^character of such genus, than upon its resemblance to some 

 single species or group of species of the genus in some one or two 

 points of small importance, or of no importance at all, generically, 



* Die Versunk. Insel Atlantis, p. 26. 



