FLOllA OF ALUM BAY. 107 



hiaintained themselves and spread over central Europe, only 

 dying out or giving way in late Miocene times. 'J'his is why the 

 Fiora of Alum Bay is of such iuimense interest and importance, 

 why its composition is so different from other Eocene floras, and 

 why it is confined to a single horizon. Misled by" its strikino- 

 facies, together with that of the flora of Monte Bolca, which 

 resembles it, and l)eing unacquainted with any other type of 

 Eocene florn, Heer set it up as a sort of test flora, determinino- 

 according to the degree in which other floras resembled it, whether 

 they should be classed as Eocene or not. Thus the floras of Mull 

 and Bovey were discarded from the Eocene, as those of Readino- 

 and Bournemouth would have been had they been adequately- 

 known at tlie time. For the same reason the representatives of 

 the Bournemouth flora on the Continent, became his type of a 

 Lower Miocene (now Oligocene) flora. 



In the present state of our knowledge no real analysis of the 

 Alum Bay flora is possible. It is remarkable for the absence of 

 any well authenticated ferns, except the pinnae of a still some- 

 what doubtful Marattia. AncBmia sjihcretacea, Sap., has been 

 recorded only as Asplenium Martinsi by Heer. As it is common 

 in the Reading Beds and again in the Bournemouth Beds and could 

 evidently support a iiigh temperature, its occurrence would not be 

 extraordinary in the Lower Bagshot Beds, but requires confirma- 

 tion. Chrysodium lanzceanum, Visiani, which abounds in the corre- 

 sponding pipe-clays of Studland, has also been recorded, probably 

 erroneously, from Alum Bay. Of Gymnosperms the Cupressites 

 elegans of our former edition has been transferred to the srenus 

 Podocaiyiis. Two specimens have revealed traces of fruit, but of 

 too indistinct a chai-acter to be very reliable. The foliage greatly 

 resembles that of Glyptostrobus which occurs plentifidly in the 

 Reading Beds beneath and the Bournemouth Beds above. There 

 appear to be no other Coniferje in the flora. Of Monocotyledons 

 none whatever are determinable unless it be a very doubtful and 

 unique orbicular leaf something like a Smilax. Palms are repre- 

 sented by a few macerated fragments that may have come from 

 the fringe of a leaf such as Sahal, and Reeds by almost equally 

 unsatisfactory fragments of sword-shaped leaves. The Dicoty- 

 ledons are probably between 40 and 50 in number, of which 

 almost all the most characteristic are absolutely confined to the 

 Lower Bagshot horizon in this country. A dwarf leaf of a 

 similar Aralia was once found in the highest Woolwich beds at 

 Lewisham, and twice the Dryandra {Comptonia) acutiloha has 

 been found in a small patch of pipe-clay low down in the Bourne- 

 mouth beds, on the last occasion in the presence of that dis- 

 tinguished palteobotanist M. de Saporta. Some of the most 

 ordinary types of leaves look as if they may be common to other 

 formations, but no importance attaches to them, and with the ex- 

 ceptions just alluded to no strikingly well-marked leaf of either 

 the Woolwich, Reading, or Bournemouth series is known to be 

 common to the Alum Bay flora. The wealth, greater than is 

 supposed, of leguminous plants is one of its chief characteristics, 

 and next in order, are the large leaves ascribed to Ficus. The 



