420 fiEtoRT— 1887. 



number to have a diminishing temperature, though the presence of Salal 

 major in the Bembridge marls, and of Nelumbium higher up, negatives 

 the idea that the climate had down to that period made any near approach 

 to temperate. 



The only one of the fruits met with in great abundance that is perhaps 

 common to Bovey is, as already mentioned, the Carpolithes, or Folli- 

 culites Wehsteri ; but in the first place it is doubtful whether the species is 

 actually the same, aud in the second it is characteristic of the Bembridge 

 marls and not of the Hamstead beds proper, and ranges downward into 

 the Lower Headon. It is a small ovate fruit, Plate III., figs. 21-27, slightly 

 curved and more or less flattened on two sides. The integument is deeply 

 furrowed or corrugated, except over the base, which is broad and smooth, 

 with a depressed scar in the centre, around which it is slightly puckered. 

 It dehisces longitudinally, and the thick, leathery, or woody, separated 

 valves are found in layers nearly half an inch deep, forming uninterrupted 

 sheets, which probably have an immense horizontal extent. A few of the 

 unopened fruits contain a smooth hollow cast in pyrites, keeled on one 

 side, slio-htly recurved, with a small scar, and truncated at one end and 

 pointed °at the other (figs. 28, 29). It is more likely that this is a 

 mere infiltrated mould of the interior than that it represents the seed. 

 In the majority of the closed fruits there is nothing but a small white 

 membranous sac (figs. 22, 25, 27) in startling preservation, but in these 

 cases the fruits are compressed and may have been abortive. There is 

 no further evidence to show whether they are cryptogamous,^ and indeed 

 they cannot yet be assigned positively to any living family or even order — 

 a fact to be regretted the more, as they ai-e widely distributed and have 

 been frequently described. We cannot help being more and more struck 

 with the fact that although resemblances can always be found between 

 livino- and fossil leaves, so that the several fragments can be fitted with 

 the name of an existing genus, very few Tertiary fruits indeed can be 

 assigned to existing genera, particularly when their structure is well pre- 

 served.2 It thus appears that at Bovey there were no leaves found that 

 presented any difficulty, and the dicotyledonous forms, however fragmen- 

 tary, were referred to twenty-one species under various existing genera, 

 such as Quercits, Dryandra, Eucaly]otus, &c. But the remarkable feature 

 of that flora, in which, as at Hamstead, fruits were unusually numerous, 

 is that not a single fruit or seed belonging to any one of the genera 

 represented by leaves has yet been found. The fruits supposed to be de- 

 terminable were three species of Nyssa, two of Vitis, two of Anona, one 

 each ofNymphcea and Gardenia, and there are seven indeterminable ones, 

 called Carpolithes. Nothing could place in a stronger light the doubt 

 attaching to the determinations come to in the case of the Bovey Flora. 



In addition to the Bembridge marls. Folliculites reappears sparingly 

 in company with a deeply interesting fruit in the Lower Headon. There 

 is a band about a foot thick near the top of that formation, both at 

 Hordwell and in the Isle of Wight, which is black with so-called ' seeds,' 

 an inch of the matrix appearing to contain some hundreds of them. It is 

 in reality a minute asymmetric, echinated fruit, appearing to be bearded at 

 both ends, and formed of a large and a small valve (pi. III., fig. 30). When 



' Described by Sir J. D. Hooker, Quart. Journ. Geol. Soc, vol. xi., p. 566, as 

 Sporangia. See notes to description of plates. 



- See Bowerbank's Fossil Fruits of the London Clay, or Von Mueller's Vegetable 

 Fossils of the Auriferom Drifts. 



