700 square miles, of which not more than a bare fifth or sixth is covered by the highest sandy bed with which the fossil trees are associated, and which from its incoherent character has everywhere suffered to the greatest extent from the action of denudation. There can be no doubt that the entire group formerly extended as an uninterrupted deposit far below the latitude of Rangoon, though the highest member of the group with its associated fossiltrunks does not extend down now in force nearer than 130 miles or thereabouts to that town, or not south of the Tounguyo nulla. The exact termination to the south, however, of this fossil-wood bed is rendered very obscure, by its merging, so to speak, in the debris which has resulted from the waste of the group, and beneath which it sinks and is lost sight of That it formerly extended much further south is rendered certain (and perhaps the occur rence, in situ, of patches beneath the newer accumulations at the present time is also indicated) by the occurrence of large pieces scattered about within the area of the detrital beds above mentioned, of a size such as to preclude the idea of transport from a distance; as, for instance, between the Okhan and Thonsay streams, where a log of not less than four feet in length is embedded in a mass of confused detritus fully 65 miles south of the spot I have assumed as the southerly limit of the group containing the fossil wood in situ. Smaller pieces of fossil wood are found much nearer Rangoon and in cuttings in the neighbourhood. These pieces on my first visit to Rangoon, and before I entertained any suspicion of the connexion of the beds at Rangoon and those containing the silicified wood, I was inclined to regard as brought to the spot by human agency, as the Burmese are fond of surrounding their religious buildings with posts of this wood "Engin chouk," but I am now convinced that such is not the case, but that the pieces in question are derived either from the wasted and missing upper beds or from the lower ones of the group still remaining, which, as I shall show, contain the same fossil-wood, though sparingly and never in the same sized pieces as the upper or emphatically the fossil-wood bed of the province. Thus the fossilwood in the Prome district occurs in two distinct formations and under very different conditions, viz., in the form of entire trunks in situ or fragmentary pieces, but little rolled, and in well worn or polished pieces, some of large size, but more frequently as pebbles, which form a conspicuous ingredient in the recent gravels.

Below I give, in descending order, a table of the main divisions into which the miocene beds east of the Irawadi may be divided, the upper three of which constitute the fossilwood group of which I am now treating

(a).-Sand-in parts gravelly and conglomeratic-characterised by the profusion of concretions of peroxide of iron associated with it; fossils, trunks of silicified exogenous wood and locally mammalian bones. In the subordinate beds of conglomerate, rolled fragments of wood as above, silicified, (that is, mineralized subsequently to their entombment), mammalian and reptilian bones and teeth of cartilaginous fish.

(b).--Fine silty clay with a few small pebbles mixed with sand in strings here and there the whole very fine and homogeneous and devoid of fossils.

(c).-A mixed assemblage of shales, sand, and conglomerates, the last very subordinate, partaking much of the characters of beds a-b; a little of the concretionary peroxide of iron. Fossils, rolled wood silicified; mammalian and reptilian bones and cartilaginous fish teeth. Towards the base, the beds contain marine shells, and pass into those of the next group.

Pegu Group.

(d).-An enormous succession of sandstones and shales of unknown thickness and not usually fossiliferous. Particular beds, however, contain fossils in profusion :

(e. g., d-1.-Hard sandstone with corals (Cladocera).

d-2.-Blue Kama clay, highly fossiliferous.

d-3-Cytherea Promensis bed or Prome sandstone,

and numerous others which cannot be specified till their fossil contents have been more especially examined).

(a).-This bed which we may fairly suppose to have been once co-extensive with the rest of the group is now greatly diminished in area by denudation, which its mineral character even more perhaps than its position at the top of the same has tended to encourage, so that even within the area where it is at present best preserved it by no means constitutes the entire surface, being everywhere deeply scored through to the underlying beds below. The surface is everywhere protected by a gravelly layer composed of small quartz pebbles and ferruginous concretions derived from pebbly strings and irregular courses of conglomerate dispersed through the sand, which readily washing away leaves the residual layer in question at top; to the protection afforded by which against further waste, the existence of what still remains of this incoherent bed is largely due. This surface layer is of variable thickness, its development being, to some extent, a measure of the denudation this group has undergone. On the surface and impacted in it at different depths where it is very thick lie logs of silicified wood of all sizes from a foot or so to trunks of 40 and 50 feet, not entire, but jointed up into pieces of various lengths through spontaneous fracture, probably brought about by their own weight, and irregular subsidence during the removal of the friable matrix wherein they were originally encased. Though, as a rule, these large logs occur as described in a gravelly debris, they sometimes occur relatively to the incoherent sand so as to leave no doubt of its being the bed wherein they were originally deposited, and on which they may be sometimes seen apparently in situ, as between Thanat-ua and Kiungee, and not only in this bed but in the beds beneath it, the same fossil-wood occurs, though in smaller pieces, and much less abundantly. The larger logs are quite unrolled, but the smaller pieces are often rounded by transport, though never to the extent seen in the pieces of fossil-wood contained in the recent gravels. When this sand rises into hills, the sides are invariably steep, and not unfrequently scarped, exposing a clean vertical section of sand with its crust of gravel at top. This sand weathers into curious pinnacles wherever an isolated stone, shell, stick, leaf, or other foreign body has afforded shelter from the direct impact of rain, and the incoherent rock all round washing away eventually leaves the protecting substance perched on a slender pinnacle of sand, which recalls the similar phenomenon of the "earth-pillars of Botten" figured by Sir C. Lyell in the 10th Edition of his "Principles."

In color this sand is greyish, very fine and uniform, and with only a certain admixture of impalpable argillaceous matter forming, where exposed to traffic, a fine dust, or, in the beds of streams where the argillaceous portion has been removed by water, a clean silver sand very fatiguing to travel over. Though the sand I am describing unquestionably contains silicified wood, yet it seems probable from the great abundance of large trees strewed over the surface, that they existed more plentifully in that topmost portion which has almost disappeared through denudation leaving only these bulky memorials behind it, than elsewhere. The structure of the wood has been to a considerable extent obliterated by decay before its mineralization was effected, and all that can be definitely said of it is that the wood is exogenous and not a conifer. I have remarked but one species in Prome, though the Burmese, from trivial distinctions in color and weathering, affect to recognise the modern Enjin (Hopea suava) and the Thiya (Shorea obtusa), an identification of course quite illusory. This wood nowhere exhibits any traces of marine action as might have been anticipated had it floated about till water-logged in a brackish or purely salt estuary, and hence it may be inferred with considerable probability that it floated about in a freshwater sea, or chain of lakes fed by a sluggish stream, till it sank where it became ultimately silicified. It must at the same time be remembered that the wood found in beds containing marine fossils is also free from perforations, but these are small pieces, much rolled prior to their entombment and probably under conditions on some sub-marine bank unfavorable to the presence of either Pholas or Teredo. In some pieces of fossil wood I have noticed minute tubular cavities (perforations ?) about 02 or less in diameter, which might have been produced by some insect whilst the trunk was still standing, but such cases are rare. Associated with this sand, and forming sometimes irregular beds in it, or more frequently lenticular courses, now thickening, now thinning out, occur some hard sandstones, sometimes very fine grained, at others a pebbly grit, or even coarse conglomerate. No regular position in the sand can be assigned to these subordinate layers, but the fine hard sandstone often occupies a high position in the deposit, whilst a coarse conglomerate is not unfrequently met with towards its base. Both sandstone and conglomerate are usually richly charged with shark's teeth of small size (Lamna), the conglomerate being usually ossiferous as

well, though throughout the deposit the occurrence of bones is irregular, capricious, and local.

A good section displaying the relation of these grit-courses to the less coherent rock around them, is seen in the hills about three miles east of Shuebandor, or about fifteen miles east of Alán-mio. The hills are here steeply cut in the bed I am describing, the surface being covered with the usual gravel, with abundance of silicified wood and ferruginous concretions, the former completely blocking many of the deep gullies cut on the hill side. Strewed about here may likewise be seen numerous lustrous fragments of iron slag, from the native furnaces once scattered over these hills. The sand here presents its usual incoherent, typical character, but a compact sandstone, passing into a coarse grit in patches, is somewhat freely developed in irregular lenticular courses in it. On the surface of the incoherent sand at this spot, and evidently weathered out of it, I picked up a fragment of the lower jaw of a deer, and from the immediate vicinity I collected mammalian bones, mostly ill preserved and fragmentary, shark vertebræ and teeth, and chelonian plates (Colossochelys and two species of Emys). In the great slabs of grit lying about amidst the debris of the wasting sand which enveloped them shark's teeth were plentiful, accompanying mammalian bones and fragments of wood, many of which had been perfectly rounded by attrition Another before they were embedded. These pieces of wood are, however, not common. locality where bones are still more abundant is one-half mile north-east of Talok, or fourteen miles north-east of Thaiet-mio on the east bank of a stream not marked on the map. Bones are here far from scarce, but friable and ill preserved. They occur both in the incoherent sand and also in the coarse grit and conglomerate associated with it, together with shark's teeth and small pieces of wood. At this spot there is a good deal of coarse conglomerate, and in accordance with the indications afforded by these coarser beds we find the bones of a larger size, and many of these much rolled and abraded before they were finally embedded. Here I got a fragment of the lower jaw of an elephant, together with fragmentary portions of the limb bones of that animal, all imperfect either from original violence or subsequent decay, the former cause certainly operating in some instances.

I may here remark that the bones found in this group (within the area I am now concerned with) are not all in the same mineral condition. The majority are somewhat imperfectly mineralized and consequently decay very readily when bared to the atmosphere by the wasting of the surrounding rock, and this I am convinced is the reason of so few bones being found on the surface, even in spots where the rock is seen to contain them somewhat plentifully. A few fragments may here and there remain, but most of the bones noticed by me were so tender, that it was clear that a short exposure to atmospheric action would reduce them to crumbling masses, which would break up and leave scarcely a trace behind them. A bone is, however, here and there found in the water courses well mineralized and calculated to defy atmospheric action, but the scarcity of these fragments attests that such is not the usual condition of bones in these beds. That these well mineralized bones are derived from the same beds as the more friable ones is undeniable. The best mineralized bone perhaps met with was the part of a deer's jaw above mentioned, and this most certainly was derived from the soft incoherent sand whereon I picked it up. The astragalus of a ruminant (Cervine?) found also by me during a former season, was in like manner an isolated example of well preserved bone, though being found in a small stream its parent bed was not demonstrable. In Upper Burmah well mineralized bones are probably more common to judge by those which have been at various times collected there, and the difference is merely the result of different conditions at the time of their deposition, such as we might expect to prevail, and depending probably on the irregular access or supply of silica in solution. That the supply of this silica must have been at some period abundant is testified by the enormous amount of silicified trunks everywhere met with, but the horizon of these is certainly higher than that at which the bones in question occur, and although small pieces of silicified wood occur commingled with the bones, it does not therefore follow that the same abundant effusion of silica took place at the time of their deposition as subsequently occurred when whole forests were silicified, and this I should be inclined to regard as the true explanation of this condition of most of the bones in this sand, viz., an insufficient supply of silica in solution.

As a rule, it is not, however, in the sand but in the coarser and more conglomeratic beds that the bones seem mostly to occur, of which a good instance is seen midway between Omouk and Lema, some 19 miles east-south-east from Thaiet-mio. Here a great bed of

conglomerate is seen dipping 30° south-east in which I noticed the tusk of a small elephant, but too friable to be extracted from its hard matrix, together with other bones, all in a poor state, and more or less injured by rolling about on a coarse shingle before their final consolidation.

Next to the presence of silicified wood, a remarkable development of concretionary peroxide of iron seems to characterise the sand I am describing. The ore occurs occasionally as a thin band, up to perhaps a thickness of three inches, breaking up or jointing into rhomboidal concretionary masses of different sizes and shapes. More usually the ore occurs in the form of variously shaped concretions from one to four inches in length, though occasionally even larger. These concretions are found in both the sand and conglomerate, to which last when numerously developed they impart a peculiar varnished look, which might sometimes be almost styled (but for the technicality of the term) viscous or slaggy. The more usual shape of these concretions is flattish oval or amygdaloidal, but they occur spherical, cuboidal, cylindrical, with both flat and hemispherical ends, discoidal and any intermediate form, but always symmetrically proportioned, and the result of a segregative action or process in the clayey and ferruginous components of the bed when in a plastic condition. Of whatever shape however, their structure is extremely uniform, consisting of an external crust of concentric layers of brown hæmatite surrounding a kernel of pure white or yellowish clay, lying loose and shrunken in the interior.

Externally these nodular concretions are roughened from the adhesion of the sand enveloping them, but this rough crust scales off readily, leaving their surface perfectly smooth. Internally they often present a blistered appearance from the mammillary crystallization of Limonite which lines them, becoming on exposure to the atmosphere and rain lustrous and varnished. Where the bed has been of too harsh a character to permit the regular segregation of the ore, it is found lining sinuous cavities in the coarse matrix, leaving flat, approximating walls, evidently produced by shrinkage, which gives such portions a very peculiar aspect and one which simulates a viscous condition. In some places even a botryoidal structure is induced where the rock is less coarse." **

The thickness of this upper sand cannot be closely estimated, but 40 feet is probably more than the average thickness of what now remains of it.

(b).-Below the last described sand occurs a deposit of very uniform character composed of pale silty clay which passes upwards into the overlying sand. This silty clay is very fine, thin bedded and homogeneous, with merely a few strings of sand here and there, and an occasional small pebble in the sand. It is everywhere seen at the base of the last bed into which it seems to pass, though their respective characters constitute a good means of demarcation between them. It is entirely devoid, as far as observation goes, of organic remains. A good section of this silty clay is seen south of Thanat-ua, between Alán-mio and Kiungalé, but the bed presents no special point of interest.

It is also largely exposed in section 14 miles east of Talok on ascending out of the stream (previously noticed as unmarked in the map), but it merely presents the same uniform character and absence of fossils, which distinguish it elsewhere. Where the upper sands have been completely denuded so as to leave exposed a large area of this bed, an undulating country is the result, possessing a marked character. The surface of the country does not there greatly differ in appearance from that seen within the area of the alluvium, and it would not

* Under the Burmese rule this ore was extensively smelted, but no furnaces are now anywhere at work in the district. Remains of furnaces which were merely rectangular kilns, cut in the firm alluvial clay of some steep bank, which gave easy access at top for replenishing ore and fuel, and below for withdrawing the products, are numerous about Shuebandor, Kiungalé, and Yebor, together with slag-heaps, sometimes of no inconsiderable dimensions. Throughout the area of these upper sands, however, slag may be found here and there scattered about, as the iron-workers shifted their scene of operations from spot to spot, wherever charcoal and ore was for the time most plentiful. The works must have in many cases been conducted in the dry season only, as the hearths of some furnaces still standing open into the beds of streams which during rain would certainly have found an entrance to them. The blowing apparatus was probably the effective vertical cylinder bellows formed of large bamboos still in use in the district by blacksmiths, but the oldest inhabitant could give me no particulars of the manufacture, as none of the class of iron-smelters now remain in the district. The introduction of English iron and steel has doubtless been the main cause of the abolition of this branch of industry, aided by the harsh and injurious system of the Burmese officials during the early struggles with the British, but in some places it was alleged that the ironworkers had fled the country to avoid being forcibly transported to Calcutta to make iron for the terrible foreigners. This may seem very absurd, but those who know the ingrained credulity and ignorance of Asiatics will be inclined to give some weight to the reason stated, though it is probable that this fear, strongly as it may once have operated, is no longer felt, though the state of the market and the price of iron now ruling in Pegu prevents the resuscitation of the trade.

be easy in a limited space to discriminate the clay in question from the ordinary alluvial clay of the valley. Where, however, freely exposed, it presents much the appearance of a regur,' save in color, which is a pale yellowish-gray, quite devoid of any tinge of red which the alluvial clay generally possesses, and equally so of the dusky carbonaceous hue of a regur. From some peculiarity in its composition or hygrometric qualities it in dry weather opens out in great cracks, and is always covered with a sparse crop of stunted grass in separate tufts, and a tree jungle of a peculiar aspect from the dwarfed character of the trees composing it, present among which are the Toukkian (Terminalia macrocarpa), Te, (Diospyros. sp.), and the "Shábiu" of the Burmese (Phyllanthus emblica).* The country around Laidi comprising the doab between the Pade and Myo-hla streams is composed of this clay with sparing remnants here and there of the upper sands. It is largely exposed, too, in the broad valley about Lepaláh (Let-pan-hla) and between that village and Chouk-soung ("stone fang"). Towards the mouth of the Myo-hla stream near Toukkian-daing, (Htouk-kyun-deing,) this clay forms the open country and is dug for making pottery. It might here be readily mistaken for the alluvial clay of the valley, but for the occurrence here and there strewed over it of small pieces of silicified wood derived from the denuded sands which once covered it. The thickness of this bed I cannot estimate, but I should not place it under 40 feet; how much more cannot be determined.

(c).-Below the last described clay, a group of beds occurs of rather varied character, resembling, to some extent, the beds both above and below it. It contains, though sparingly, the same description of fossil wood as the sands at the top of the group, and some of its beds present characters very similar to portions of those beds; whilst towards its base, it appears to pass insensibly into the lower group characterised by marine fossils. It is, however, generally very devoid of organic remains, though, as a convenient lower horizon to it, I have taken a sandstone which is generally recognisable where the junction is clear, by a few organic remains not very well preserved, among which a coral (Cladocora) is most characteristic, which we may regard as the highest member of the lower group.

A section of these beds is seen in the Kini-choung (Kyeeneech) above Mogoung, which may be taken as illustrating their general character, and some portions so resemble the ossiferous sands and gravels of the upper beds that I searched confidently, though in vain, among them for like fossils.

(Descending).

Very false-bedded pebbly sandstone
Harsh sandstone, rather irregular

...

20 0 2 30

(a few feet).

Compact yellowish silt with a central band of kidney-shaped nodules 1 to 2 feet
in diameter

74 9

This section, though not a thick one, will illustrate the general character of the upper portion of this division (c). The silty shale much resembles the shale in division b, whilst the sands equally recall the uppermost sands, (a.) Close on the horizon of the above section

The clay above described and the sandy beds of the same group, respectively, offer good instances of the connection of particular soils with particular kinds of vegetation. So generally does this hold good in Pegu that in some instances it affords a good empirical criterion of the geological formation beneath. In the area of the fossilwood sands, the most prominent tree is the Eng (Dipterocarpus grandiflora), and this tree so commonly affects a sandy soil that the Burmese call such soils, whether within the limits of the fossil-wood sand proper or the zone of detrital accumulations skirting the hills, "Engdaing," or the tract of the Eng tree, and though, of course, Eng trees are found on other descriptions of soil, yet it is on this sandy belt that the Eng flourishes most vigorously from probably being there less competed with by other trees, well fitted as it for a sandy soil. The "Thiya" (Shorea obtusa, Wall,) the "Kanyin" (Dipterocarpus alata, Wall,) and the "Engyin" (Hopea suava, Wall,) equally affect the sandy "Engdaing," though not in sufficient numbers to characterize the forest. On the other hand, these trees abhor the clay described above and are most miserably dwarfed on it. The Toukkian (Terminalia macrocarpa), though dwarfed, seems to answer best on this clay, but from some cause or other it does not seem favorable to vegetation. I think this must be due rather to its hygrometric properties, than to any injurious ingredient in it, and that if artificially irrigated, it would give better promise to the cultivator than the densely wooded sands to which it offers so unpleasant a contrast.

Bamboos are not usually much developed on the Engdaing, and a striking demarcation is not unfrequently seen where the Engdaing meets the boundary of the older beds on which bamboos flourish with great luxuriance. The Burmese are fully alive to this fact, and if an enquiry is made regarding a village, say, if it stand within the Engdaing, will answer it negatively, "it is among the bamboos," an expression quite equivalent in their minds to saying it is not on the Engdaing where bamboos are rare and never are the characteristic vegetation,