Sept 22, 1 88 1 J 



NATURE 



495 



difficulty in working the cave is the removal of the debris to pre- 

 vent its interfering with further work. We therefore put up a 

 windlass over the eastern entrance and cleared a way for barrows 

 through the tahi> below. Beneath the talus the black earth, in 

 which remains had been previously found, was seen, and many 

 articles of Roman age were taken from it. Chamber III. was 

 marked into parallel, and these into squares. In the centre we 

 sank a shaft and passed through the following layers: — (i) 

 Romano- Biitish layer, a black earth with pottery, ornaments, 

 &c., and numerous boi es, usually from i to 2 inches thick; 

 (2) hardiih stalagmite, abuut 6 inches thick, in one place con- 

 taining the bones of a dog or small wolf ; (3) soft stalagmite 4 

 inches thick ; (4) hardish stalaginiteb inches thick ; (5) soft stalag- 

 mite 2 feet 6 inches thick ; (6) stiff broivn clay w ith large angular 

 blocks of limestone fallen from the roof firmly imbedded in it. 

 This layer was 8 feet deep, as far as we saw it. The last two 

 feet are laminated and contain smaller blocks. At the depth of 

 about 12 feet fn m the surface we came upon part of the solid 

 limestone floor or side of the cave, sloping steeply downwards. 

 There were no indications of a change in the nature of the 

 deposit at the junction with the limestone, .and the clay appears 

 to extend much deeper than the level at present reached. Thus 

 below the stalai^mite purely mechanical deposits succeed, and no 

 limestone blocks are found above this horizon, although the 

 stalagmite has been removed over a large ] art of the floor of the 

 chamber. No traces of a fauna have been as yet found below 

 the first hardish stalagmite ; indeed all the deposits passed through 

 below the stalagmite indicate the former presence of a still lake 

 in which the great thickness of clay slowly accumulated. Further 

 work was stopped by the heavy rain which flooded the shaft dug 

 in the clay. It is interesting to note that the former condition 

 of Chamber II. is identical with the present state of the third 

 chamber in the preponderance of mechanical over chemical de- 

 posits. The change from mechanical to chemical deposits was 

 probably produced by a change from accumulation in still water 

 to accumulation in running water. Possibly also the absence of 

 blocks fallen from the roof in the stalagmite may be due to the 

 bicarbonate of lime contained in the water which percilated 

 through the roof, cementing together the limestone blocks. The 

 absence of this cement when the clay was dep :)sited may be due 

 to the ab ence of solvent power in the water which then perco- 

 lated through the roof. For the carbon dioxide would not be 

 evolved fro n a soil deficient in organic matter, as the soil cover- 

 ing the Yorkshire hills for a period long after the Glacial period 

 must have been. The author expresses his best thanks to Mr. 

 J. R. Tennant of Kildwick Hall, Leeds, .and to Mr. J. R. Eddy 

 of Carleton, Skipton, who gave, on behalf of the Duke of 

 Devonshire, the permi-sion to work the cave, and further aided 

 with kind help and advice all through the work. 



On Asteromilta Readi, a New Speeies of Coral from the Oligo- 

 cene of Brockenhurst, by Prof. P. Martin Duncan, F. R.S. — The 

 author described the characters of this coral, placing it in the 

 genera establi-hed to include certain corals from the West Indies, 

 and some dredged up in the Carribean Sea by Count Pourtales. 

 He referred to the genus Madrepora, which lives in tw-enty to 

 twenty-five fathoms, 74° Fahr. temperature, reef-building coral, 

 or on banks in a turbulent sea. The specimens are generally 

 rolled, but some are absolutely perfect, and clearly give the 

 history of the phyical conditions of the close of the Eocene 

 period in the south of England, which tben reseii.bled the climate 

 of the Bermudas. 



On the Formation of Coal, by E. Welhered, F.G..S., F.C.S.— 

 The author considers (i) that coal was not formed from vegeta- 

 tion of the Lepidodenoroidtype, and that therefore the Stigmaria 

 found in the underclays are not the roots of the vegetation which 

 gave rise to the coal ; {2) that the varieties of coal and the 

 change which sometimes lakes place in one and the same seam 

 are not due to metamorphism, nor are they dependent upon the 

 contorted state of the surrounding strata, 1 ut arise from the greater 

 or less chemical decomposition of the vegetable mass, influenced 

 by the circumstances under which it was submerged. On the 

 land grew the vegetation of the period, represented by the 

 Lepidodendrons, feigillaria, Calamites, &c. As the land sank 

 and the waters encroached, the land vegetation was gradually 

 washed a^ ay, but the roots remained in many cases, and those 

 which offered the greatest rcistance to decay are the ones pre- 

 served in a fossil state — hence the occurrence of Stigmaria. As 

 the waters advanced the ground would become swampy, and 

 then we might expect to see spring up reeds, mosses, and other 

 vegetation suitable to the changed condition ; it is to vegetation 



of this kind that the author ascribes the formation of coal. 

 With a view of ascertaining whether the chemical composition 

 of the beds which overlie a seam of coal which has changed 

 from bituminous to anthr.acite also changed, the Welsh " nine- 

 feet" seam was selected, which near Cardiflf is semi-bituminous, 

 and at Aberdare becomes anthracite. Specimens of the over- 

 lying strata were selected from the two districts at each foot 

 above the coal for five feet ; these were analysed, and it was 

 found that the beds from near Cardiff were considerably more 

 argillaceous and, as a whole, less ferruginous than those at 

 Aberdare. 



On the Palctozoic Rockt of North Devon and West Somerset, 

 by W. A. E. Ussher, F.G.S., Geological Survey of Engknd 

 and Wales. — The classification adopted is as follows : — 



LOWER 



DEVONIAN 



MIDDLE 

 DEVONIAN 



Foreland 

 Grits 



i Red and purplish grits, fine- 

 \ grained, and in places sili- 



UPPER 

 DEVONIAN 



Lynton 

 Bebs 



Hangman 

 Grits 



Ilfracombe 

 Slates pass- 

 ing INTO 

 MORTE 



Slates 



PiCKWELL 



Down Beds 



Baggy Beds 



Pilton Bedsc 



pla 

 [ ceous. 



( Grey, even-bedded, and jointed 



5 grits, grey schists, and schis- 



J tose grits with films of calca- 



( reous matter. 



/ Coarse white quartzose, r.;d- 



speckled grit, in and upon red 



and grey rather fine-grained 



1 grits associated with shaly and 



slaty beds. 



Grey and silvery slates and 



shales with arenaceous films, 



J and impersistent bands of 



I limestone pas-ing into pale 



greenish unf ossiliferous quartz- 



» ose slates. 



j Indian-red slates upon red, 

 I green, and grey gi'its, with 

 , local purple slate basement- 

 I beds passing into the Morte 

 \ slates. 



Green slates with Lingqiila; 



I brown micaceous grits with 



Cuculltca, positions of these 



horizons apparently reversed 



near Wiveliscombe. 



r Bluish and greenish grey argilla- 



I ceous slates, with occasional 



thin films of limestone and 



masses of grit (as at Braunton, 



, &c.). 



The Foreland Grits occupy an area (superficial) of thirty square 

 miles, extending from Countesbury to Dunster. The Hangman 

 Grits form the range which includes Dunkery Beacon, also the 

 whole norlhem part of the Quantocks. Their relations to the 

 Ilfracombe Slates are much complicated by faults around Croydon 

 Hill and on the Quantocks ; and the prevalence of grits in the 

 Ilfracombe series, whilst indicative of lithological assimilation, 

 makes the boundary rather indefinite. 



On the Characters of the " Lansdcnvn Encrinite" (Millericriuus 

 Pratii, Gray, sp.), by P. Herbert Carpenter, M.A. — The 

 " Lansdown Encrinite " is a species of Millerierimis (M. Pratii, 

 Gray, sp. = Apiocrinus obeoniciis, Goldfuss) from the Great 

 Oolite on the top of Lansdown, near Bath. It is remarkable 

 for the very great variation in the characters of its stem and 

 calyx. The former may reach 50 mm. in length, and consist of 

 seventy di-coidal joints ; or there may be less th,in ten joints, the 

 lowest of which is rounded off below, and its central canal clo ed 

 up. Various intermediate conditions may occur between these 

 two extremes, while in some specimens there may be only two 

 to four stem-joints ; and in one case the w hole stem is repre- 

 sented by a .slightly convex imperforate plate on w hich the liasels 

 rest. This specimen, taken by itself, would be naturally re- 

 garded as a Comatula of advanced age, in which the cinhus- 

 sockets had disappeared from the centre dorsal just as they do in 

 the recent Actinometra fukcsii. The general appearance of the 

 calyx is very similar to that of Pentacriniis W'y^ille-Thomsoni 

 from the North Atlantic. But it is remarkable for the number 

 of small intercalated pieces which it may contain. The basels 

 are frequently separated from one another, or from the radials, 

 by minute plates which, while regularly developed all round the 

 calyx in some specimens, are entirely absent in others. The 

 nearest allies of M. Pratii seem to be M. Munsterianiis, var. 



