Astrocenia clavellata, Terq. et Piette, is found in the Luxembourg Lower Lias, but the zones above that of Ammonites angulatus in the Lias do not present, as yet, any species. The species is represented in the Oolites, and became extinct in the Falunian.

Cyathocania is a new genus, which I have suggested and published in order to admit forms which, had they been furnished with a columella, would have been classified as Astrocœniæ. There is a species in the zone of Ammonites Bucklandi. Some of the species are mimetic of the Astroconia. The following is a scheme of the genus.

Cyathocania with

Cyathocœnia.

branching, having costæ

incrusting, no cost, conenchyma granular the corallum.... flat large costa, and a deep calice

globular, no costa, cœnenchyma plain..

C. dendroidea.
C. incrustans.
C. costata.

C. globosa.

The gradation of structure in the genera just passed under our notice becomes more and more evident as such forms as those included under the genus Cyathocenia are studied. In the early stage Thecosmilia cannot be distinguished from Montlivaltia; but gemmation from the calice, from the calicular wall, or from the wall ensued, or fissiparous division occurred, as the corals grew. There was an evident tendency in Montlivaltia Walliæ, Duncan, for instance, to reproduce by calicular gemmation; but in Oppelismilia distinct calicular budding occurred. Under these circumstances the genetic relations of the three genera Thecosmilia, Montlivaltia, and Oppelismilia are of the closest.

Now in bush-shaped Thecosmiliæ union often occurs between a bud from the wall and the parent stem. A section transverse to the line of growth shows, (1) low down, two corallites with their septa, walls, and epitheca perfect; but higher up the epitheca is not seen in a section, and the walls may be (2) slightly separate, or (3) quite fused, and they then appear as one lamella between the corallites.

(1) is what is observed in Elysastræa; (2) is the Septastræan peculiarity; (3) peculiarizes Isastraa.

The origin of Latimeandræa from Isastrea has already been noticed. In Elysastraa the epitheca and one wall become absorbed at the calicular margin. In Cyathoconia the epitheca between the walls becomes cœnenchymal, and variously ornamented; whilst in Astroconia the same thing occurs besides the growth of a columella.

The following grouping of the genera is made with a view to assert that they had genetic relations with Montlivaltia, and that some Cainozoic types revert to more ancient.

a. An evident reversion to Elysastrea in a recent genus.

b. Solenastrea is a case of reversion to an ancestral Thecosmilio-Elysastræa type in the later Neozoic ages.

c. Has great probability of being a case of atavism with much modification of Thecosmilia and Solenastraa.

The Montlivaltia of the zone of Ammonites angulatus are remarkable for their septal regularity, the amount of dissepimental endotheca, the usually rudimentary condition of the true wall, and the development of the strong and compensating endotheca. These characteristics are observed in the St.-Cassian Montlivaltic, and in those which are found in the strata intervening between the Ammonites-angulatus and the St.-Cassian beds. These structural peculiarities, in a genus whose later Jurassic species have a perfect hexameral arrangement, a perfect wall, and moderate endotheca and epitheca, indicate the descent of the Montlivaltice from a Palæozoic stock. In Montlivaltia Murchisoni, Duncan, the wall and epitheca are united perfectly into one structure with the intercostal spaces, just as the septa of some simple Palæozoic corals are continuous, not with costa, but with the intercostal spaces or their analogues.

M. parasitica is remarkable for its septal number; and M. simplex has distant and curved septa.

M. papillata, Duncan, M. Hiberniæ, Duncan, and M. Haimei, Chap. et Dew., are closely allied species; they are broad-based, pedunculate, short, and turbinate, and vary greatly. The last-named species ranges probably over the whole area of the zone of Ammonites angulatus.

Closely allied to Montlivaltia and Thecosmilia is the new genus Oppelismilia. Its Palæozoic aspect is distinct, the multiseptal and non-cyclical calice, the calicular budding, and the strong epitheca all refer it to bygone types.

These corals, from the Lias beneath the zone where Gryphoa incurva and Ammonites Bucklandi are abundant, indicate that, like the succeeding formations of the Chalk and the Oolite, the Lias was very coralliferous. Nothing marks the progress of paleontology more strongly than the ability of making this statement from well-ascertained data; for within a very few years the Lias was considered so muddy a deposit as to be obnoxious to coral life.

Now, with a great fauna, part of it indicating reef conditions and the rest moderately deep water, the Lower Lias will assume as great an importance to the zoophytologist as the Eocene. The Liassic coral-fauna reflects the Palæozoic as the Eocene foreshadows the Recent coral-fauna. Unfortunately the paucity of our information respecting the earliest Secondary coral-fauna, that of the Lower Trias, is so great that the Liassic species are still greatly removed from the original types.

Corals from the Zone of Ammonites Bucklandi (bisulcatus).

Corals are not numerous as regards their species in this zone, and the com

monest species of the zone of Ammonites angulatus have not been found in any of its strata.

It is probable that Thecosmilia Martini, E. de From., which in France ranges from the beds containing Ammonites Moreanus into those in which Ammonites bisulcatus is found, has a corresponding vertical range in England. Thecosmilia Michelini, Terq. et Piette, appears to be present in the zone of Ammonites bisulcatus; but as yet only casts of its specimens, which resemble those of the species from Abbot's Wood in the zone of Ammonites angulatus, have been found. These casts, and some of Thecosmilia Martini, have been assigned to the genus Cladophyllia, but without sufficient reason. Thecosmilia is a large genus, and of course the species present individuals of all sizes; so that to give to small cylindroid Thecosmilice the generic appellation of Cladophyllia is unreasonable. In fact this last genus is but a subgenus of Thecosmilia at the best.

List of Species from this Zone.

1. Montlivaltia Guettardi, Blainville.
2. Septastræa Eveshami, Duncan.
3. Lepidophyllia Stricklandi, Duncan.
4. Isastræa endothecata, Duncan.

6.

5. Isastræa insignis, Duncan.
Stricklandi, Duncan.
7. Cyathocœnia globosa, Duncan.

Septastraa Eveshami has very irregular calices; and when the wall has been worn away between them, a groove is seen indicating that separation of the corallites which I have already noticed to obtain in the Elysastræœ. The species is rather abnormal; for although fissiparity is common, still there is a disposition to serial increase.

The genus Isastrea has three well-marked species in this zone, and they are very distinct from those of the zone of Ammonites angulatus. In Isastraa endothecata the depth of the calices, the extraordinary development of the endotheca, and the great number and the irregularity of the septa are differential. Isastraa Stricklandi also has a great development of endotheca; for large plates of it cross the corallites, and shut in the calicular fossa below, acting perfectly like tabulæ, just as in Cyathaphora. The septa are few in number; and no cyclical arrangement is to be noticed.

Isastra insignis belongs to a section of the genus which comprises I. Henocquei, Ed. & H., from the Hettangian, I. polygonalis, Muschelkalk, and I. Lonsdalei, Ed. & H., from the Inferior Oolite.

A new genus, Lepidophyllia, has a species in this zone, and a very fine one in the zone of Ammonites Jamesoni: it is an interesting form, and presents some Rugose characteristics, such as a repeated calicular gemmation and an epithecate wall.

The only Montlivaltia I have seen from the zone of Ammonites Bucklandi has a lower horizon on the continent. Having thus a very considerable vertical and geographical range, the species is, of course, very variable, and many local varieties have been found, which are separated with difficulty from Montlivaltia Haimei. These flat multiseptate Montlivaltice are very characteristic of the Lower Lias. They have a representative in the zone of Ammonites obtusus, in the form of M. patula, Duncan, whose dentate septa are wonderful. Such septa began then to be the fashion; for in the next zone the Montlivaltia are famous for their grandly ornamented dentations.

Corals from the Zone of Ammonites raricostatus.

The Montlivaltia from Fenny Compton, Honeybourne, and Cheltenham belong to several species, and two of these are singularly polymorphic. Shape has not much to do with the specific diagnosis of some recent simple

corals; and it is necessary to assert this in collecting under one fossil species corals of very different external forms. The Montlivaltia from the zone of Ammonites raricostatus are common, and their mineral condition has been preservative of the minutest details: even the granulations on the minute septal dentations are preserved.

Dr. Wright collected and described a very remarkable series of corals from the Hippopodium and coral-beds of Marle Hill, Cheltenham, Honeybourne, and Fenny Compton, naming them Thecocyathus rugosus. The assemblage of forms thus named contains very varied specimens, the external shape especially being rarely alike in two or three instances. A careful examination of sections of most of the forms enabled me to place them all in the genus Montlivaltia. The absence of pali and the presence of short endothecal dissepiments proved that they could not belong to the genus Thecocyathus. But the general Montlivaltian characteristics have also the paleozoic peculiarities already noticed in considering the Montlivaltia of the zone of Ammonites angulatus. Montlivaltia rugosa, Wright, sp., will therefore take the place of Thecocyathus rugosus, Wright, MS.

Montlivaltia mucronata, Duncan, is a polymorphic species, remarkable for its elegance and ornamentation; some of its specimens are amongst the most beautiful of the Madreporaria. The study of a large collection enabled me to place some very different-looking forms in the same species, the intermediate varieties having been in my possession.

There is a decided affinity between these Montlivaltic and M. Stuchburyi, Ed. & H., of the Inferior Oolite. Moreover the M. nummiformis, Duncan, of this zone is related, if structural affinity be of value, to M. lens, E. & H., Inf. Oolite. Montlivaltia radiata, Duncan, is a very abnormal species, and retains the quadrate septal arrangement, which is fully represented in many Liassic species, but which is so characteristic of many Palæozoic forms. It must be remembered that such strange structural peculiarities in later forms may arise from atavism.

List of Corals from the Zones of the Lower Lias above the Zone of
Ammonites Bucklandi.

Montlivaltia patula, Duncan.

rugosa, Wright, sp.
mucronata, Duncan.

Montlivaltia nummiformis, Duncan.
radiata, Duncan.

There are then twelve species in the Lower Lias above the zone of Ammonites angulatus, five of which are above the zone of Ammonites Bucklandi. It needs no care to decide that the fauna of the zone of Ammonites angulatus has little affinity with that of the other zones.

Corals from the Middle Lias.

1. Lepidophyllia Hebridensis, Duncan.

2. Montlivaltia Victoria, Duncan.

The first species is from the island of Pabba, and was collected by Dr. Wright; it forms a bed there, and was doubtless a rapid grower.

The genus has already been slightly noticed; its calicular gemmation and the growth of epitheca on the free wall of the corallites, where they grow higher than their neighbours, refer to an Elysastræan, Thecosmilian, and Septastræan series.

A great number of specimens of all sizes of a very polymorphic Montlivaltia have been found on the surface of the fields at Chemington, near Skipton, and in a watercourse or ditch section of the Middle Lias close by. Ammonites

Henleyi, A. Chiltensis, Cardinia attenuata, and C. elongata were found with the corals.

Montlivaltia Victoria, Duncan. This coral grows to a height of five inches, and may be two inches broad; it is the largest simple secondary form, and has the epithecal wall so peculiar to the Liassic Montlivaltia. Its septal number is very great, and the endotheca is highly developed. It is very variable in shape.

There are some fragmentary corals in the Marlstone, but their genera are doubtful; and the cast of a Montlivaltia was found by Mr. Charles Moore at Wells, but I cannot determine the species.

Corals from the Upper Lias.

The only species is that which was found years since, and which was described by MM. Milne-Edwards and Jules Haime, Thecocyathus Moorei, Ed. & H.

Total number of Species of Corals from the British Liassic Strata."

The descriptions and drawings of sixty-six of these species are in my Monograph of the Liassic Corals,' 1867, 1868, Pal. Soc.

The Thecocyathus Moorei, Ed. & H., was described and drawn in the 'Monog. Oolitic Corals,' by Milne-Edwards and Jules Haime, Pal. Soc.

Report of a Committee appointed to investigate Animal Substances with the Spectroscope. By E. RAY LANKESTER.

DURING the year attempts have been made to obtain a supply of Siphonostoma or Sabellæ for the purpose of investigating the derivatives of the body described by me last year as chlorocruorine; at present a sufficient supply has not been obtained. The absorption-spectrum of chlorocruorine from Sabella, however, has been carefully observed and recorded. The Sponge-chlorophyll has been investigated with the object of determining which of the two green and two yellow bodies, spoken of by Professor Stokes as being present in plants, is present in the sponge; and some interesting results appear likely to be obtained when the history of plant-chlorophyll is more fully known.

The feathers of twenty-two species of birds, mostly red, green, or blue, have been examined for absorption-spectra; none was obtained; but Prof. Church has discovered a red matter containing copper in the feathers of the Turacou ; and to this body he gives the name turacin. The spectrum of this substance I have carefully examined and recorded. As stated by Prof. Church, it gives two absorption-bands, when in the feather, close to those of hæmoglobin, but readily separable from them, and by no means indicating anything like identity in the bodies, as Prof. Church appears to have thought.

A scheme with the chief solar lines and Sorby's standard interference lines

1868.

K