161 PROCPEDINGS or THE GEOLOGICAI4 SOCIETY. 



beds of an earlier epoch presenting as many species as are preserved 

 in this, ferns especially predominate. It must, however, be borne in 

 mind, that in the temperate floras of many parts of the northern 

 hemisphere, neither ferns nor gymnosperms are usually associated in 

 any great numbers with large-leaved dicotyledonous trees. In Japan, 

 I believe that Coniferae and large-leaved Dicotyledons are associated, 

 as also in some parts of America ; in England ferns generally accom- 

 pany similar forms of foliage to the Reading fossils ; whereas in Si- 

 beria, where broad-leaved dicotyledonous trees are abundant, ferns are 

 extremely rare ; as is also the case, I believe, in many parts of North 

 America. 



There is no foHage resembling that of grasses, sedges, or rushes, 

 among the Reading specimens, nor any appearance of organs of 

 fructification. 



With regard to the individual fossils of these beds ; all, except 

 figs. 22 and 23, are decidedly Dicotyledonous and Exogenous, and 

 none present structural or physiological characters of importance, 

 either in texture, form, or nervation. Figs. 22 and 23, from having 

 parallel veins, may be assumed to be monocotyledons ; of these, 22 

 is too anomalous-looking for me to hazard even a conjecture as to the 

 appearance or nature of the plant to which it belonged ; while there 

 are few natural orders amongst Monocotyledons to which 23 might 

 not be referred. 



Though the leaves preserved in the Reading beds are all of the 

 very commonest forms in the vegetable kingdom (of Dicotyledonous 

 plants), I do not find that they exactly resemble those of any living 

 English species ; and indeed even were the resemblance so close that 

 I could not distinguish them from existing forms, I should not con- 

 sider myself warranted in drawing any conclusions therefrom ; be- 

 cause, in the first place, the normal or typical form of leaf in any 

 species can seldorn be decided by one specimen or at one epoch of 

 growth ; secondly, because very similar leaves may belong to very 

 different species ; thirdly, because the top, base, margin, and stalk of 

 a leaf are all absolutely essential for identification of the species to 

 which it belongs, and these are not all present in any of the Reading 

 specimens ; and fourthly, because in these, as in all fossil leaves, the 

 important characters of texture, pubescence, and colour are necessarily 

 obliterated. 



It would be very easy to produce from an herbarium leaves so 

 similar to 1, 2, 6, 12, 13, 14, 18, &c., as to deceive the inexperienced 

 into instituting crude afl[inities ; but after a very careful comparison 

 of these fossil leaves with those of willows, poplars, oaks, maples, 

 Myricse, Rhamni, and such familiar genera as must suggest them- 

 selves to every one, I find that while I cannot advance beyond plau- 

 sible suggestions, nor give better reasons for such affinities than those 

 presented by outline, I can adduce a copious list of far less familiar 

 genera belonging to widely diff'erent natural orders, to which there 

 are as good grounds to refer these leaves, as to the genera I have 

 enumerated. It must indeed be evident to any one acquainted with 

 the real value, in a systematic point of \iew, of characters derived 



