worn as we ascend. The sandstones here and there afford evidence of their derivation from granite rocks, while beds of coarse waterworn debris occur at intervals, the contained boulders being nearly if not quite as large as below.

Close to the village of Ning-kong jow the beds change, the sandstones become more frequent, close grained, hard and extremly compact; the intermediate beds become also finer and in some places might be called consolidated mud stones. Occasionally these latter begin to assume tufacious characteristics and seem as if poured out over the surfaces of the sandstone beds. The colours are various shades of brown approaching to red. The approach from the underlying beds is so gradual that it is difficult to define the junction but the Ning-kong jow flag-stones may be taken as from 800 to 1000 feet in thickness. In the centre of the series the sandstones occur in beds of from 2 to 8 feet in thickness dipping regularly to the N.E. at an angle of about 73° to 9°. These afford admirable building stones, and are extensively quarried for doorposts, lintels tablets &c. The stones are readily removed by wedges, the cleavege being perfect in the direction of the bedding. Stones upwards of 20 feet in length and a foot and a half in thickness are readily procured in this way, and are well adapted for the ordinary trabeated bridges of the country, for which purpose they are shipped away in large quantities. The stones when worked exhibit a fine surface either with or across the grain and are excessively durable.

Ascending the Ning-kong jow valley in a direction to the north of west these flag stones are seen capping the hills to the right hand and gradually increasing in altitude till at the head of the valley they attain a height of about 1500 feet. They are very conspicuous as their superior toughness and durability have preserved them from the denudation which has extensively eroded the lower rocks of the river.

Above the flag-stones conglomerates again occur, but they now begin to become greyish, and assume a more tufacious aspect. The continued boulders are smaller and are not confined to the palæozic rocks, but contain fragments of the lower Du-bu-du rocks, showing that denudation had commenced with more or less oscillations of level. Crossing a range of low hills towards N.N.E. on the Tszechi branch of the river, the upper members of the series are seen in what may be called the Da-ying tufas. The character of the rocks has here completely altered, and in place of sandstones we find grey or greenish grey tufas, the debris for the most part of trachytic volcanic products, but in places assuming a reddish tinge, as if

dolerite alternated with trachyte. In the trachytic magma frequently occur fragments of the paleozoic rocks as well as of the lower beds of the series; the bedding has not become confused, and the rock has assumed a secondary cleavage independent of the bedding. In places the texture is so fine that the rock affords an excellent and durable building stone capable of showing the finest detail under the chisel; and like the Ning-kong jow flag-stones is extensively worked, but principally for carved brackets, panels, and other fancy work. For the most part the structure is closer and the contained pebbles, often of the lower argillaceous shales, render it unsound for such purposes. As however the cleavage spoken of above is very marked, the rock splitting readily into flagstones from 3 to 5 inches thick, it is extensively used for paving throughout the district as well as at Shanghai.

These rocks extend in a W.N.W. direction as far as Hangchow, where about the Lui-fung pagoda they may be seen cropping out in the low hills bounding the Sihu. They seem here to pass into red sand stones, apparently similar to the red sandstones of the Nanking district. To the N.E. they reach the plain of Yu-yao, and are cut off by the paleozoic rocks which reappear above the city of Tszechi. It is many years since I visited this district, and at the time I had not had the opportunity of studying the lie of the lower Ningkong jow series, so that the entire was a sealed book to me. Speaking only from memory I cannot venture then on more than the merest outline. The rocks however must be of considerable thickness, probably exceeding that of the Du-bu-du beds, so that the entire to the commencement of the red sandstone cannot be under 10,000 feet.

As to the age of this extensive series we have at the moment only geological and lithological data to form an opinion. So far as I know the entire of the system has never yielded a single fossil. This was of course to be expected in the lower conglomerates, the conditions of whose deposit indicated considerable meteorological disturbance. The upper beds of these and the Ning-kong jow sandstones, deposits in comparatively settled water, might have been expected to yield some signs of life. Except however a few obscure molluscan or worm tracks on the ripple marked surfaces of the beds I have hitherto failed to find any trace of an orgainzed body. We are thus left without the only sure ground from which the age of the formation can be inferred. The evidence of position is likewise vague. The rocks overlie the palæozic series of central China, and were deposited after it had been uptilted and altered, but even this leaves a wide interval, and we are forced to reason from analogy.

The position of the rocks flanking the palæozoic ranges of Chehkiang, the heterogeneous character of their contents, the occurrence of the coarse conglomerates and boulder beds, and the subsequent subaerial denudation remind us forcibly of the Siwalik beds of northern India, and the apparent absence of fossils strengthens the resemblance. It is only here and there in the Indian beds that fossiliferous beds have been discovered, though where found the bones have been discovered heaped up as in a charnal house, indicating apparently, as do likewise the boulder beds, the occurrence at times of wild cataclysms. The divisions of the beds are roughly similar, though the thick deposits of volcanic tufas do not occur in the Sub-Himalayic beds; also the fact that the upper beds are mostly formed of debis from the lower indicating considerable local disturbance during the deposit of the series.

As above stated the rocks have undergone considerable aqueous denudation; the courses of this are however simple, and we can follow them in the present configuration of the country. We miss the complicated systems of denudation and re-denudation, to coin for the nonce a word, which marks the older formations. The lines are sharply cut, not blurred, and the mark of the graving tool is everywhere apparent. This gives a newness of aspect to the rocks, which is increased by the low angle and regularity of the dip, from 4 to 8°. The latter is, it may be stated, no test of age; as comparatively modern rocks, the Miocens of the Alps and Himalayas for instance are constantly found contorted, vertically bedded or even inverted. Still the facies of the whole seems comparatively recent, and we seem justified fully in referring them to Tertiary times. In a case of this sort what may be called geologic instinct come into play, and though the instinct may be at times sadly at fault, the practical geologist knows from experience that in the majority of cases the forecast comes approximately if not absolutely true. European geologists who are apt to refer cases of difficulty somewhat too freely to glacial phenomena, have seen in the very similar boulder beds of the Alps and the Sub-Himayalas the traces of a Miocene glacial epoch. I am by no means however disposed to refer every case of the occurrence of boulders, even of large size, to the action of ice, and in this instance the characteristic traces of glaciation are absent. Coarse conglomerates and boulder beds do however indicate the existence of considerable disturbing causes during their deposition; and the geological evidence is accumulating that the Miocene was a period of extreme disturbance. provisionally therefore in the absence of fossils, and to afford a standpoint for reference I am disposed to class the Ning-kong jow

series as of Miocene age, probably in the upper Da-ying tufas reaching to lower Pliocene.

Such provisional classification is however tantalizing to the geologist, and as many of the contributers to the Recorder are well acquainted with the district, they might do a service to the science by enquiring amongst their Chinese friends as to their knowledge of fossil finds. The pliocene beds of Szechuen yield a rich mammalian fauna, the fossils being well known as lung-kuh-dragons bones. Although it is not likely that such exist in quantity in the district in question, even an isolated specimen might serve to determine the geological relation of the beds. Molluscan casts or the remains of fishes or plants, though less decisive than the mammals, might still throw much light on the subject, and I am loth to believe that the entire series is unfossiliferous.

אן

EVOLUTION OF FINAL K AND t out of p, anD OF T OUT OF K.

BY REV. J. EDKINS, D.D.

N Goddard's Tiech'eu vocabulary may be noticed a great readiness to drop final p or change it into k or t. Thus to pierce should be c'hap but we find ch'a, Z bap, deficient is hwat, k'iap, weak is k'iak, whereas in the kwang yün it has p final.

It may be remarked here that the kwang yün has final k in the words,,,, all which or their phonetics are also found with final p.

The last of these

word for "to eat"

pronounced with k final at Shanghai in the is in Tiech'eu k'iet. In both cases there has

been an evolution from p, as we conclude from the phonetics, J.

It is breathing in or hip.

In the syllable k'iet we find,

Solat has come from lik and sat

but these are kok in kwy. from sak.

In sek we find which in the old dictionary kwang yün is shap or shak. Here k has been evolved from p. The word to leak issiak or siap, and here also k is from p, so also t comes from siet, to tie.

pinor

The words,, called tit are properly spelt with k final.

In the kwang yün we find among k finals the following phonetics in p佳羽,奎,夾,嗇,皂,舞,兹,止,乇,四,幸,羽,立,卒, ,, and others. They may be found with p in pages 42 to 50 in the Ju sheng volume of that work.

In the same way if we look for p phonetics among words ending in t in the kwang yün we find . This phonetic requires us since its original final is p to regard kot the grass cloth plant, kot, to cut, and kie to finish, exhaust, as all ending in p. So with #,,, 毦,刃,内,熟,岁,止,兒,茶,臬,隶, and others, They indicate that an extensive migration has taken place from p to t. An example occurs in the Amoy pronunciation of hwat. The dictionary final is p and so it is at Tiech'eu, but Amoy speech has adopted it.

The final p has been best retained in Kiang-si province and the old k in Shanghai and at Fucheu in both which cities final p and t are quite lost.

Confirmation of this doctrine of the evolution of k and t from p is to be found in old forms of phoneties. Thus is, but has in its upper part. So also, below the net at the top, has the same form for . In the same combination occurs. The old form for we thus obtain is very like the old form of which is and of which is . We may regard this form therefore as originally having the force dap and kap. The evolution of k from p and t is not confined to the final. The initial letter is also subject to the same law. Thus pau to embrace is pok. But to embrace is kwo or kwok, the environs of a city. Should any one say these must be seperate roots, it may be replied, that it is easier to change a letter than to create a new root, Ease of origination is a principle that must not be lost sight of when determining what are true roots.

In etymology it is important to know the true origin of several groups of words which have sprung luxuriantly from roots ending in p. Lip or dip, to stand, has originated shu, tree, chu, an individual tree, Eshu, upright, † chu, pillar, #chu to support, ‡ chu, tablet. All these words mean upright and have lost p, but they have it is most likely first changed p to k and then dropped k.

Who has not looked curiously at a character like, c'hi, in tonic dictionaries gi, which means c'ha, and also coincides with chï branch is sense, while yet it differs in initial, having g instead of aspirated or unaspirated t (ch)? the fact is that all these words began with t'ap or dap and they are ultimately identical with + hip, mean_ ng a cross, the common numeral shi, ten. The guttural g is evolved from the tooth letter d which appears as t', zh, ch, or sh.