278 Count Gaston de Saporta on the Temperature 



first time in Tertiary formations,, and even in the fossil state, I 

 dwell upon this determination. This genus, with only a single 

 exception, includes only arborescent species : in this case more 

 especially it denotes a truly tropical type, because the species 

 from Sezanne are closely allied to the most tropical forms of the 

 genus in both hemispheres, whilst they differ from those which 

 advance furthest towards the south, such as A. pruiiiata, Kaulf. 

 (Chili), and A, «Ms/r«/2s (Tasmania) . 



The repeated presence of palms in the Soissonais sandstones, 

 and that of a frond at Sezanne too conformable to the bifid type 

 of Carludovica not to denote the persistence of the type of the 

 Pandanese, constitute a combination of tropical forms which is 

 further increased by a Cissus formed upon the model of C. fer- 

 ruginea, Poir., indica, Roxb., tomentosa, Lam., repens, Thw., 

 adnata^ Wall., and capejisis, Thunb. (all perfectly tropical forms), 

 by a Sterculia, several Tiliacea and LaurinecBj &c. 



. The second of our horizons presents, of tropical types, — 



Palms: Flabellariaj Sternb. — F. parisiensis, Brong. (Calcaire 



grossier.) 

 Sabalites, Sap. — S. sp. [Flahellaria maxima^ Ung., Brong.). 



(Oise, Crisolle.) 

 Pandane^ : Nipadites, Bow. — Several species observed in the 



London Clay, in Belgium, and in the Calcaire grossier 



of Paris. 

 Sapindace^ : Cupanioides, Bow. — Several species in the London 



Clay. 

 TiLiACEiE ? : ApeibopsiSj Heer {Cucumites, Bow.). London Clay. 



The fossil fruits described by Bowerbank under the name of 

 Nipadites, and of which that author was able to examine the 

 structure in the pyritous specimens of Sheppey, differ in no im- 

 portant character from those of Nipa, a curious genus which 

 seems to form the connexion between the Palms and the Pan- 

 danese : it is reduced at present to a single species, Nipa fruti- 

 canSy which inhabits the banks of the Ganges and the marshy 

 parts of Java. The Nipadites, first observed in the London Clay 

 and afterwards in Belgium, have lately been met with in the 

 Parisian Calcaire grossier. The working of the mound of the 

 Trocadero has enabled a great number of impressions to be col- 

 lected in a sandy clay bed, in which there are also observed 

 monocotyledonous leaves analogous to those of certain Carludo- 

 viccEj but entire, which should, perhaps, be combined with these 

 organs. 



The Cupanioides of the London Clay reveal the existence of 



