fearon ert) ass at a rer, Se Se ee, eu snees ater r 5 of ies arg eta * Le es mit PERS ey ae Seen = Ke campeon Katto Meee 7 eS a eng Sie ead : oe ter « 5 Seine ee eae ae ae Petre eto Se y= . ou. A — oe ie ie IT ic : Aen A mie, Sa hh gee one? a Se ~ Lia ee es er. Gorad PALAONTOGRAPHICAL SOCIETY. INSTITUTED MDCCCXLVIL. VOLUME FOR 1880. MDCOCLXXX. Praag oN uy iene 4 ins ‘a ae A MONOGRAPH OF THE BRITISH EOCENE FLORA. BY JOHN STARKIE GARDNER, F.G.S., M.G.S. Francs, &c., AND CONSTANTIN BARON ETTINGSHAUSEN, Pua.D., PROFESSOR OF BOTANY IN THE UNIVERSITY OF GRATZ, BTC. ETC. PART II. Pore rs. Pagzs 39—58; Prares VI—XI. LONDON: PRINTED FOR THE PALHONTOGRAPHICAL SOCIETY. 1880. PRINTED BY J. E. ADLARD, BARTHOLOMEW CLOSE. EOCENE FERNS. 39 Some difference is apparent between the angles of the venation of the young pinnz attached to the plant and those drawn on PI. X, figs. 9, 10, which are more acute. The original specimen, which should be preserved in the Jermyn Street Museum, is not now to be found there ; and our identification rests, therefore, partly with Heer, who states that they are the same. It would be better, perhaps, that the species should be described under the name of GontopTEris than under that of Pancopreris, as at page 38. Goniopreris stiriaca (Unger). PoLypopirTes stir1acus, Unger. Chloris Protogzea, p. 121, pl. xxxvi. 1847. Lastrma stiriaca, Heer. Flor. Tert. Helv., vol. i, p. 31, pl. vi and vii; vol. iii, p. 151, pl. exlin. 1859, — HELVETICA, Heer. Ibid., p. 33, pl. vi, fig. 2; vol. ili, p. 151, pl. exliii, figs. 2—5. 1859. GoniopTerRis sTiriaca, 4. Braun. Zeitschrift Geolog. Gesellsch., vol. iv, p.556. 1852. — — Schimper. Pal. Végetale, vol. i, p. 547. 1869. Lasrr@a strriaca, Heer. Flora of Bovey, p. 28, pl. v, figs. 12—15. 1861. — — Gaudin et Strozeit. Contributions 4 la Flore Foss. Italienne, vol. ii, p. 32, pl. i, fig. 2. 1859. PHEGOPTERIS STIRIACA, Ettingsh. Flora of Bilin, p. 16, pl. ii, f. 16—18. 1866. G. fronde pinnata, pinnis linearibus, prelongis, inferioribus grosse crenatis serratisve, supertoribus argute serratis vel serrulutis ; nervatione Goniopteridis Aspidit, nervo primario valido prominente, recto, nervis secundariis sub angulis 50—65° orientibus, tenuibus, sub- rectis vel paullo arcuatis, nervis tertiaris in pinnis inferioribus plerumgue 6—7, in pinnis supervoribus plerumque 4—5, curvatis, subparallelis, angulo acuto egredientibus. Soris rotundatis biservatis. Middle Bagshot, Bovey Tracey. This Fern has not, we believe, been met with in England elsewhere than at Bovey Tracey, where it was formerly found in abundance. ‘The pinne are usually attached to the rachis, and the sori are frequently preserved. The stems and circinate vernation placed with Pecopteris lignitum by Heer’ belong, in all probability, to this species.” We have no hesitation in uniting Lastrea helvetica, Heer, with the present species. The former was first founded upon most imperfect fragments, and separated upon the supposition that it was more closely and sharply serrated, and possessed fewer ternary veins. ‘The much more perfect specimen figured in the third volume of the ‘ Flor. Tert. 1 «Flora of Bovey,’ p. 31. 2 See description of Osmunda lignitum, p.,50. 40 BRITISH EOCENE FLORA. Helvet.’ shows that the separation cannot be maintained ; for in the drawings there are in places seven veins in Z. helvetica to five in LZ. stiriaca; while comparisons of the tracings of Heer’s figures,' here reproduced, show that Z. stiriaca is really the more strongly toothed. It is impossible, in fact, to compare the figure of Z. stiréaca, vol. i, pl. vin, with that of Z. helvetica, vol. iii. pl. exliii, fig. 2, without recognising that the latter is the upper part of the frond of the former. The species was founded upon Fre. 16.—L, helvetica. ‘Flor. Tert. Fre. 17.—L. stiriaca. Ibid., vol.i, Fic. 18.—G. stiriaca(?). «Fl. foss. Aret. ? Helv.,’ vol. iii, pl. exliii, fig. 2 (part). pl. viii, fig. 4. pl. xlv, fig. 7. : insufficient material, and should have been annulled when the more perfect specimens were figured. Specimens of Z. dalmatica which we have received present in like manner no distinguishing specific characters, and should not remain separated. In the ‘ Flora Fossilis Arctica,’ p. 87, pl. xlv, fig. 7, we find a portion of a pinna figured under this name from Atanekerdluk, in Greenland. It possesses, however, according to Heer’s description, only three ternary veins on each side of the pinnule ; while in the description of the Swiss form six to seven were essential to the species, and Fie. 19.—G. stiriaca. ‘Flora of Fic. 20.—G. stiriaca, Ibid., fig. 13. Fie. 21.—G. stiriaca. Enlarged by Bovey,’ pl. v, fig. 14 (part). Mr. Fitch. specimens possessing fewer are separated. He therefore seems to modify his specific characters to include it, and in modifying them thus, leaves them sufficiently open to 1 Figure 16 shows the strongest denticulation present in the drawing of the fine specimen of L, helve- tica in the third volume. p EOCENE FERNS. 4] admit several other supposed species. It appears from the figure that the pinna is more tapering, and therefore shorter, than those of G. stiriaca, and all the veins more sinuous. The indistinct specimen from Cape Lyell has the exact outline of this species. G. stiriaca is supposed by Unger and Heer to resemble G. prolifera’ “ of tropical America,” which Fern, however, it is well known, is a native of all the tropics except that one. The venation is not uncommon among Ferns, and is met with in Acrostichum and Nephrodium, as well as in Goniopteris, but the arrangement of the sori is more charac- teristic of the latter. In the fossil state it has been found in France, Germany, Italy, &c., and seems to characterise principally the Middle and Upper Eocenes, and Oligocene. It appears to have been of much larger growth than G. Bundurii, and possessed more numerous and simply curved, instead of sinuous or angulated, veins. The original specimens are too indistinct to be satisfactorily refigured, and we have, therefore, repro- duced fig. 19 and part of fig. 20 from pl. v, ‘Flora of Bovey,’ and an unpublished ¢, 21) by Mr. Fitch, of Kew, who has placed his original sketches for that enlargement (fig work at our disposal. Woopwarpia ? venosa, Zt. and Gard. Plate X, fig. 5, 5a. WV. fronde pinnata, lacinis anguste lanceolatis, margine integerrimis ; nervatione Dictyopteridis simplicis exappendiculate ; nervo primario recto, prominente, nervis secundariis brevissimis tenuissimis angulo acuto egredientibus ; maculis Dictyopterrdis pluriseriatis, inequalibus. Middle Bagshot, Bournemouth. Although but one small and not very distinct fragment has been met with, the character of the venation is very apparent. The frond may have either been simply pinnate or bipinnate. ‘There are a number of species of Péerts,as P.patens, P. decurrens, P. denticulata, P. Woodwardioides, &c., with anastomosing venation and closely similar ; we cannot, therefore, be certain that it is really a Woodwardia. We have, however, placed it in that genus, because a very beautiful and undoubted species of Woodwardia has been met with in the Eocene of Monte Promina, &c., and the Aquitanian stage of Switzerland. This form, VW. Roessneriana, Unger, resembles W. radicans, and differs from ours in the possession of a row of free veins between the areolations and the margin. Our species would fall into the group of Lornseria, Presl, of which W. areolata, a native of the United States, from Massachusetts to Florida, is the best known, and, with one exception, the only existing type. 1 Ett., ‘Farnkrauter,’ pl. cix and ex. 2 Flora of Bovey,’ p. 28. A2 BRITISH EOCENE FLORA. ADIANTUM APALOPHYLLUM, Saporta. Plate X, fig. 1, la. ADIANTUM APALOPHYLLUM, Saporta. Flora of Sezanne, Mem. Soc. Géol. France, t. vill, p. 313, pl. 1, fig. 1. 1865—1868. A. pinnulis obovatis subtrapeziformibus, margine superiore sinuatis, nervatione Cyclo- pteridis simplicis e@qualis; nervis primaris tenuissimis aequalibus dichotomis, soris apicalibus subconfluentibus. Middle Bagshot, Bournemouth. Small Adiantoid pinne had been met with among the interlaced fronds and stems of Gleichenia at Bournemouth (see p. 44), to which we considered they might have belonged, but, as similar pinne have since been found unassociated with G/eichenia and preserving distinct remains of sori, there is no longer any doubt that a species of Maidenhair, not much larger than Adiantum capillus-Veneris, formed part of our Middle-Eocene flora. As we see no reason to consider them distinct specifically from those found by Count Saporta at Sézanne, which only seem to differ in the absence of sori, we have placed them, with his concurrence, in that species. ‘The segments seem to be dimidiate, as the line of fruit is only present on one side. ‘This species differs strikingly from the group of A. capillus-Veneris in the linear, imstead of obversely reniform, arrangement of the sori. Detached segments are occasionally met with among dicotyledonous leaves in beds west of Bournemouth Pier. Hewarpia rucia, Hit. and Gard. Plate II, fig. 5; Plate VII, fig. 2; Plate X, fig. 6; and Plate XI. H. fronde membranacea simplicr trregulariter inciso-lobata, subrotunda, longe petiolata, basi subtruncata, lobis lanceolatis integerrimis, undulatis vel sublobatis, margine soriferis, soris confluentibus ; nervatione Dictyopteridis simplicis, nervis tenuissimis, maculis angustis lanceolatis. Middle Bagshot, Bournemouth. The palmate form, rectangular base, and long stipes show the specimen figured (Pl. XI, fig. 1) to have been an entire frond, and not a large terminal pimna. So much of the frond is preserved that little is left to the imagination in restoring it. It was nearly equal-sided, palmate, somewhat irregular in form, bluntly six-lobed more than half-way down, each lobe being again once subdivided. It measured six inches EOCENE FERNS. A3 along the base, which forms almost a right angle with the stipes, and at least four inches in length. he texture is exceptionally membranous, the sori, where the lobes overlap each other in places, being distinctly traceable through it; this character removes it from Péerzs, in which genus it might otherwise possibly have been placed. The stipes was glabrous, rigid, slender, and must have been at least three times longer than the portion preserved. The veins are very fine and distinct, and anastomose copiously, but become less conspicuously reticulated towards the margin, as seen in the enlargements, Pl. X, fig. 6 @; Pl. XI, fig. 4. The sori form long linear patches upon both edges, but do not extend to the extremities of the lobes. The specimen, PI. II, fig. 5, is probably a small barren frond of the same species. Only the outline is preserved, which is suborbicular, with margin entire. The slender stipes, membranous texture, and marginal sori are characters only com- bined in Adiantum and Lindsaya. In the latter genus L. sagittata possesses simple fronds and is not very inferior in size; but it is in the section of Adiantum with anasto- mosing veins, Hewardia of J. Smith, that we find the nearest resemblance to it. H. Wulsoni, although not possessing perfectly simple fronds, sometimes has its pinne almost all united in a single trilobed terminal pinna, approaching the fossil species in dimen- sions. ven the nearest living species, however, differ very widely from it, and it thus appears to be one of the few Tertiary Ferns that have become completely extinct. We are not acquainted with any fossil species with which to compare it. We had frequently noticed fragments of a very membranous Fern with anastomosing venation, and two of these had been figured without any clue as to the form of the leaf, before the magnificent specimen on Pl. XI had come to light. Our attention was called to it by a lady, who found it in a small concretion of white clay near the base of the black clays, in the leaf-beds nearest to, and on the west side of Bournemouth Pier. The other fragments were found in the same beds. (r) Gleicheniacee. Genus. —GLBICHENIA. (Sub-genus.— MErTEWSIA.) GieicHenta Hantonensts (Wanklyn). Plate VI and Plate X, figs. 2, 3, 4. Merrensites Hantonensits and M. crenata, Wanklyn. Annals and Magazine of Natural History, vol. ii, pl. i. 1869. G. fronde gracili stipitata dichotoma ; stipite scandente ; cirrhis abbreviatis uncinatis ; AA. BRITISH EOCENE FLORA. rachi tenui; pinnis oblongis ; pinnulis approwimatis patentibus, linearibus vel lanceolato- oblongis, margine tntegerrimis vel crenatis ; apice obtusis ; nervatione Pecopteridis vere ; nervo primario e rachi angulo subacuto, rarius recto exeunte, tenui, plus minusve flexuoso ; nervis secundariis cum primario angulos acutos formantibus, abbreviatis, catadromis, furcatis, rarius simplicibus; nervis tertiartvis catadromis, arcuatis, simplicibus, rarius furcatis, enfimo lateris exterioris receptaculum punctiforme emittente. NSoris singularibus vel duobus rotundis, lobatis. Middle Bagshot, Bournemouth. The stipes and rachis are slender, and are always found associated with the remains of, and sometimes attached to the pinnules The repeatedly dichotomosing character of the stem is well seen in the specimen, PI. VI, fig. 8, in which is also shown one of the young buds produced at the forking of the larger stipes. The stipites were furnished with short, hook-like tendrils. Figs. 9 and 10 show these tendrils with remains of the rachis, and fig. 11 with the attachment to the stipes only. On this piece and at fig. 8 are spirally rolled buds. ‘The relative position of the tendrils and pinnz is shown on PI. X, fig. 2. The remains of pinnee, represented in Pl. VI, figs. 1, 2, 3, and 7, show them to have been of an elongated, almost lanceolate shape, bearing linear or lanceolate pinnules of unequal length. These are closely set, with crenate or entire margin and rounded apex. The venation is that of Pecopteris vera. ‘The midribs of the pinnules are sinuous and relatively slender; they leave the rachis, which is itself slightly flexuous, at angles of 65° to 80°, rarely either more acute or obtuse. The secondary veins leave the midrib in cata- dromous order at the same angles as those formed by pinna and rachis, and are mostly forked, rarely simple. The ternary veins are also catadromous and mostly simple. The sori occur on one or both of the exterior venules of each group of veins. ‘The sporangia were deciduous, and arranged round a small, oval, punctiform receptacle. There are visible six to eight impressions to each cluster, but only the bases of these are seen in places where the leaf has chipped off. The fossil form is distinguished from all existing Glecchenie by the possession of tendrils or hooks, indicating a more pronounced creeping or climbing habit. With this exception, however, it bears a close resemblance to two recent species, forming sections of the sub-genus Mertensia, G. pectinata and G'. dichotoma. ‘The size and form of the pinne, the branched and zigzag stipes, the buds occurring in the forks, the venation, and the sori, are all similar. The resemblance is considerably heightened by the very unequal lengths of the pinnules, even when side by side, and by the fact that only those which are much longer than the rest are regularly crenated, as if the crenations were an expression of luxuriant growth. Adiantoid bracts are sometimes developed at the forkings, and these, being found associated with the fossil Glecchenza, had been thought to indicate the presence of Adiantum. G. pectinata, which it the more particularly resembles, is a native of tropical South EOCENE FERNS. 45 America, where it is common; and G. dichotoma, according to Hooker, is of “ tropical and subtropical regions, almost universal, in the New and in the Old Worlds, Pacific Islands, and as far north as Japan.” A note by S. Hayes, in the British Museum Herbarium, informs us that the former grows in extensive patches along the line of the Panama Railway. The genus Gleichenia appears to have reached its maximum development in Europe in Pre-eocene or so-called Cretaceous times, and then to have died out or migrated, since no species have been met with in the Miocene. Saporta has found similar fragments in the gypsum of Aix,in Provence. Fragments are also met with at Aix-la-Chapelle, which closely resemble it, but are stated by De Bey, after comparing specimens, to be all more or less distinct. The distribution of this Fern in England is almost limited to a thin band, only a few inches thick, in the leaf-patch under the Highcliff Mansions at Bournemouth. In this layer it was abundant, and numerous specimens were obtained by residents, in some of whose collections they are still preserved. It was only after an interval of many years that some small fragments were found last autumn, the result of extended digging in a neighbouring patch. Its restricted distribution here should be taken into account in generalising on the absence of Glecchent@ m the Miocene of Hurope. The position in which the remains are found, overlying a marly bed, penetrated every- where by its rootlets, seems to imply that it flourished in a considerable patch in a wet clayey hollow. Mr. Wanklyn kindly lent the specimens figs. 1, 3, 4, 6, Pl. VI; Mr. Pender figs. 2 and 5; and the Woodwardian Museum figs. 2, 3, 4, Pl. X. ANEMIA suBCRETACEA (Saporta). Plates VIII and IX. ASPLENIUM SUBCRETACEUM, Saporta. Flor. Foss. Sézanne, Mém. Soc. Géol. France, 3e ser., vol. viii, p. 315, pl. xxiii, fig. 4. 1868. GymnocramMMa Hayprnu, Lesquereuz. U.S. Annual Report, 1871, p. 295 and Tertiary Flora, U. S. Geol. Survey of the Territories, vol. vil, p. 59, pl. v, figs. 1—3. 1878. A, stipite. dichotoma, fronde bi-tripinnata, pinnis ovato-oblongis, pinnulis adnato- sessilibus, subdecurrentibus lanceolatis vel lanceolato-linearibus, acuminatis, grosse dentatis vel inciso-lobatis, lobis acutis integerrimis vel crenulatis ; loborum nervatione more Cyclopteridis simplicis equalis, nervis tenuissimis valde congestis, mediano vie prominente. Middle Bagshot, Bournemouth. The materials for the determination and description of this Fern are abundant and 46 BRITISH EOCENE FLORA. exceptionally well preserved. Pl. IX, fig. 2, represents an almost perfect frond (unfor- tunately reduced in the engraving to one fourth the natural size), in which there is a double forking of the stipes and a tripinnate imb. The pinne are ovate in general out- line, with sessile pinnules, almost decurrent at their base, as in Pl. VIII, fig. 1, or more rarely, and in the lower part of the pimne only, shortly stipitate (Pl. IX, fig. 2). The pinnules vary from broadly to linear lanceolate, and are very acuminate at the apex. ‘The margins are coarsely toothed, laciniate-dentate to lobed, the lobes being acute or crenulate. The venation, magnified on Pl. VIII, figs. 3 and 4, is of the type of Cyclopteris simplex @qualis. here is either no midrib, as in the fig. 4 referred to, or it is faintly visible, as in fig. 3, and the veins are very slender and close together. No traces of fertile pinne have hitherto been met with. The repeated forkings of the stipes and the type of venation point more to the genus Anemia than to Asplenium, as dichotomy is characteristic of the Schizeacee, and similar forked branching occurs in Anemia, either in the stipes of the sterile fronds only, or in both the fertile and sterile fronds. Saporta and Heer do not agree in our determination, and the former has kindly stated at some length his reasons for differing: they are, that no known living Anemia has so lanceolated and attenuated a form, nor so coriaceous a texture ; and that each segment, whether entire or subdivided, is attached directly to the rachis as a pinnule or pinna, and — is never either decurrent or confluent. Anemia, he thinks, is adiantoid in growth, and has nothing in common with this Fern. On the other hand, he felt, when he described it from Sézanne, extremely doubtful whether it should be placed with Asplenium; and, these doubts being now further confirmed, he has suggested’ that, since it appears to be an extinct form with doubtful affinities, a new genus, perhaps allied to Zodea through T. barbara, should be formed to receive it. Dr. Stur, from an examination of the plates and a small specimen which I forwarded to him, believed it to be an Osmunda. We cannot find, however, that any existing Osmunda more closely resembles it than O. regalis, and this seems so unlike in general growth that the data for placing it in that genus appear to us insufficient. Neither are we able to consider it an Asplenium ; for, although many species of that genus resemble it somewhat in their general free growth, cutting, and venation, yet no vestiges of the sori, so abundant and persistent on the fronds of Asplenium, have ever been found in any locality upon any of the great number of specimens examined. It seems almost certain that the fertile pinnze were separate ; and, as we have seen in the similar cases of Chrysodium and Osmunda, therefore rarely, if ever, preserved. Notwithstanding that it does not agree very closely with any existing species of Anemia, the general form and dichotomy of the frond, the venation, the length and strength of the stipes, the separation of the fertile and barren pinnee, have, after much consideration, induced us to place it in the genus Anemia. In addition to the genera mentioned, it is also not unlike Z/yrsopteris and some 1 In letter. HOCENE FERNS. A7 Davallie, but no known Fern resembles it so closely as Anemia adiantifolia,’ Sw., an ihabitant of Cuba, Bahamas, Florida, and Mexico to Bahia. Saporta has kindly compared specimens from Bournemouth with those of Sézanne, and writes respecting them, “ Je mhésite pas a déclarer qu il s’agit dune seule et méme espece.” Lesquereux, to whom we are much indebted, has also compared Bourne- mouth specimens with those from the Eocene of North America, and has written, ‘ Mt LL EP LEE a <——<—Z a SIMMS NE PLATE VII. From tHE Mipp.e Bacsnot, BouRNEMOUTH. Lygodium Kaulfusst Heer). \V. jog (it Figs. 1, 3, 4, 5, 7. Pinnee of various sizes. 8. Pinna; the specimen should have been an inch longer, and only a part is engraved. 6. Portion magnified to show venation. Hewardia regia, Kitt. and Gard. 2. Fragment for comparison with Lygodium. \/- / (The above in the Gardner Collection.) PLATE VIII. From tHE Mippie Bacsnot, BouRNEMOUTH. Anemia subcretacea, Saporta. Fig. 1. Part of a large pinna. 2. A small pinna. Vv. tEgr| 3, 4. Portions enlarged from different parts of frond for comparison of venation. (The above in the Gardner Collection.) ine 3 PL A ee” Va Ze i jy | § NN. D+ Appt Z ail “AN ) ‘{f Mig nd ae poten ee An Tie ages En) YPN Ni) a) ); ys Wp Syl, ae ———= ES S SS ZZ ZE A Zig Z LEZ —= SSS S WWest &Co imp BOCHNE- FERNS. PLATE IX. From tHe Mippire Baesnot, BouRNEMOUTH. Anemia subcretacea, Saporta. Figs. 1, 8, 5. Fragments of pinna. (Gardner Collection.) 4. Parts of pinnules. (Mr. Pender’s Collection.) te 2. A large frond, reduced to one fourth the natural size. (Gardner Collection.) © © HN EP ER Ns. as PIVAD heise ) ASN nit H \ HT A Wese Newmark Coan. 09 PLATE X. From tHe Mippie BacsHot, BouRNEMOUTH. Adiantum apalophyllum, Saporta. (Gardner Collection.) \\ . Part enlarged. Gleichenia Hantonensis, Wanklyn, showing hooks and foliage attached, and sori. . Enlargement, showing two sori to each set of veins. . A sorus magnified. Pinne with sori. Pair of hooks, with fragments of the midribs of pinne attached. (Figs. 2 to 4 in Woodwardian Museum.) WVoodwardia venosa, Ett. and Gard. . Enlargement. (Gardner Collection.) Hewardia regia, Kitt. and Gard. \ . Enlargement. (Gardner Collection.) Goniopteris* Bunburii, Heer. A nearly entire pinna. . Enlargement of the venation, Venation of another pinna enlarged. (Gardner Collection.) Lygodium Kaulfussi, Heer, Fertile frond. (Mr. Baldry’s Collection.) \ From THE OLpHAVEN Beps, BromLey. Fragment of Fern, undetermined. . Enlargement. (Gardner Collection.) From tHe Woo.twicu and Reapine Beps, Counter Hit. Pteris Prestwichii, Ett. and Gard. * Termed Phegopteris on p. 38. ip. = - 2 12) : A a v = e BOCHNE FERNS. PLATE XI. From tHE Mipp.Le Baesnot, BouRNEMOUTH. Hewardia regia, Kitt. and Gard. Fig. 1. A magnificent specimen, showing pene sori, 9 Vs te G07 2, 3. Parts of smaller pinne. Vanes 4. Portion enlarged, showing anastomosing veniation. (Gardner Collection.) PLATE .X1. np. Co3 West Newmar MOCHENE FERNS. HOCENE FERNS. 59 The preceding table? is full of suggestive interest and contains new evidence bearing’ upon questions of the highest physical importance. In conclusion the principal types are recapitulated, in order to include details of distribution and relationship to other recent and fossil Ferns omitted in the specific descriptions. 1. Guercnenta Hantonensis, Wanklyn. (Page 43.) The only Gletchenia yet met with in the Hocene closely resembles G. dichotoma, the sole existing representative of a well-defined subsection of the genus. The agreement between them is striking, and extends to some minute details ; thus, the primary veins are grouped and divided identically, the sori are identical in form, and, like in G. dicho- toma, although usually situated on the upper or left hand secondary vein only, some- times occur on both the secondary veins. This duplication of the normal single sorus to each primary vein characterises De Bey and Ettingshausen’s genus Didymosorus, a genus which can hardly now be sustained since G. dichotoma has sometimes more than one sorus to each primary vein. From the very fragmentary condition of the few specimens figured by De Bey and Ettingshausen, it cannot yet be said that the possession of two sori to each primary vein is a constant character. But were the fixity of this character apparent there might now be hesitation in attaching specific value to it. ‘The stipes, rachis, and even the leaves in vernation were glabrous. Specimens of Gleichenia dichotoma from Sydney and East Australia appear less rigid and more membranous in character than those from elsewhere, and the Hocene species 1 We have but to consider the almost complete absence of any records of other herbaceous plants, the difficulty with which mountain species, which form a large proportion of fern-floras, or species with such habits as Ophioglossum, Asplenium ruta-muraria, A. Ceterach, &c., could find their way into these deposits, to realise that we are probably only acquainted with those ferns whose habits favoured their preservation. But as nearly the whole of them have been found at Bournemouth, almost on one spot, and as repeated search has brought no new forms to light, there can be little doubt but that we possess a nearly complete knowledge of the Eocene lowland forest fern-flora of at least this locality, and probably no existing area would furnish a greater variety of types under similar conditions. Although, therefore, the number is small compared to other orders of plants, we cannot help believing the fern-flora to have been extremely rich, considering, as we have said, the complete absence of genera even as cosmopolitan as Asplenium, whose habits are less favorable to preservation. The British and foreign Ferns provisionally placed together in this table were united, in some cases, from a comparison of illustrations and descriptions only, and the grouping is therefore to a certain extent but an indication of useful work still to be carried out. 10 SS eee e ar se ~ A togt GO Ne at - ie ote