Art. XL.—A Preliminary Account of the Geological Features of the Christ church Artesian Area.

By R. Speight, M.A., M.Sc., F.G.S.

attaches to this tract of country is not one of economical character alone, since an accurate examination of it may throw some light on the structure of the Canterbury Plains, and indirectly on questions of more theoretical interest, such as the cause of the Pleistocene glaciation of this country, which has lately attracted so much attention. The amount of evidence to be considered is very great, as no part of the earth's crust in New Zealand has been more thoroughly explored than this area. The number of wells already sunk extends to thousands, and of late years all well-sinkers in the transaction of their calling, in order to be able to give accurate estimates of the cost of sinking in various localities, have kept detailed records of the wells they have sunk, and made careful note of the depth of waterbearing beds, the amount of water obtained at certain levels, the nature of the strata encountered, and the thickness of the beds.

The records of various well-sinkers have been placed at the disposal of the present author, and every assistance in the way of information has been given when it has been asked for. The writer wishes especially to thank Messrs. J. Osborne and J. W. Horne for assistance in this respect. It has thus been possible to examine the records of more than five hundred wells, so that an accurate conspectus can be obtained of the whole water-bearing area.

General Structure OF THE Canterbury Plains.

The detrital deposits of the Canterbury Plains have been laid down on a basement rock of uncertain character, but there is evidence which suggests that the same formations as occur in the western part of Canterbury are continued beneath the plains to the eastward, and extend under the sea towards the Chatham Islands. The greywackes and slates of the Southern Alps outcrop again near Gebbie's Pass, on Banks Peninsula; and the Chatham Islands are formed of schists analogous to those of Westland, overlaid by Tertiary limestone and volcanic rocks, the former of which can be correlated with limestone of Miocene or Oligocene date in the main islands of New Zealand.

The only evidence of the structure beneath the sea which stretches from the Chathams to Banks Peninsula is that afforded by the collections of the steam-trawler “Nora Niven” when carrying on an experimental cruise on behalf of the New Zealand Government. As recorded in a note to a previous paper by the present author, the trawl brought up from a number of stations parallel with the coast-line, and in depths of between 20 and 40 fathoms, pieces of brown coal sometimes as large as an ordinary travelling-trunk. It is hardly credible that these could have dropped from passing steamers, seeing that brown coal is rarely if ever carried by sea, and it is impossible that they could have been brought down by rivers from the coal-seams that occasionally outcrop on their banks, since such pieces would be rapidly reduced to powder and become indistinguishable among the other detrital material. It must be concluded, therefore, that the coal has been derived from an outcrop on the edge of a submarine escarpment which runs parallel to the coast and with its beds probably dipping west. These may continue westward, and, passing under the plains, junction with the coal-measures which fringe the eastern flank of the mountainous district of Canterbury. The beds would then take the form of a syncline, slightly tilted towards the east, with its eastern wing depressed beneath sealevel. Further, they may form part of a great ge-syncline or synclinorium

extending over the whole area from the Alps to the Chatham Islands, with schists outcropping on its eastern and western visible limits.

The uncertainty of the soundings between Banks Peninsula and the Chathams renders it impossible to speak definitely of the form of the seabottom in that region, but it appears from the few that have been obtained that the sea gradually deepens for about forty miles, to the 100-fathom line, and then suddenly drops to over 1,000 fathoms, a depth which is maintained to the vicinity of the Chathams. Whether this is part of a submarine plain or a portion of a fold valley which runs in a north-easterly and southwesterly direction parallel to the coast of the North Island is at present uncertain, and it is hoped that the efforts being made in Wellington to get a line of soundings run from Lyttelton to the Chathams may be successful, as it will throw considerable light on the evolution of the main crustal features in this region.

That a syncline exists involving the Cretaceous coal-measures, and probably the overlying limestones, seems to be very reasonable, but the cause of its form is uncertain. It may be due to deep-seated movements of the earth's crust, of which we can say little at present, or it may be due to loading caused by the immense quantities of detritus poured into the sea by the great Canterbury rivers. The presence of a line of earthquake origins parallel to the coast-line of the plains suggests crustal movements along a line in the neighbourhood of the probable submarine coal-outcrops.

Source and Character of the Material out of which the Plains are Constructed.

The detrital matter out of which the plains are constructed consists chiefly of shingle with an admixture of sand and silt, the whole being formed from the weathering and disintegration of the greywackes out of which the great mass of the Canterbury mountains is formed. In the higher parts of the plains, near the base of the mountains, there is a considerable mixture of angular matter and large subangular blocks, but in the lower parts of the plains the gravel becomes much smaller. The fragments are usually from 2 in. to 4in. in diameter, and rarely exceed 8in.; they are well rounded, and seldom exhibit the flattened ovoid form produced by the incessant drag of the backwash on a beach subject to heavy seas. It cannot, however, be inferred with certainty, because the pebbles on the plains are usually of such equilateral dimensions that they must have been formed wholly by river-action, as the shape of a pebble on the beach is primarily determined by the shape of the block from which it has been derived. On examining a beach of limestone pebbles on the north side of Amuri Bluff, I was struck by the frequency of the cricket-ball size and shape, and this could only be due to the original fragments having been cuboid in form. The angular fragments produced by the disintegrating action of the frost and other agencies on the mountains of Canterbury are usually of equilateral dimensions. However, on examining the beach at Birdling's Flat, between Lake Ellesmere and the sea, it will be noticed that a very large proportion of the pebbles exhibit the true shingle form as distinct from gravel. If the plains had been formed by marine action they must have been subject on their eastern margin to attrition by the same heavy seas that sweep up the present Ninety-mile Beach, and therefore beach-shingle should form a large part of their material. As this is not the case, I think it may be inferred that

the plains have been built up on a land-surface by the aggrading action of the rivers issuing from a mountain tract, where they are plentifully supplied with sediment and flowing on a high gradient, and then depositing their load at the base of the mountains, where their transporting-power is less. Such plains occur under somewhat similar conditions at the base of the mountains in central Asia and central United States, where marine action cannot be called in to explain their approximately level character. The present contours of the plains, as shown by the detailed surveys of Haast and Doyne, emphasize their fluviatile origin since they are just those which would have been exhibited had the plains been built up by the coalescence of the fans of low gradient formed by the large rivers as they issued from the mountains.

Formation of the Eastern Part of the Plains.

There are certain factors which must be allowed for when considering the formation of the eastern portion of the plain. The rivers have been bringing down for a long period vast quantities of the detritus. On the Ninety-mile Beach this is exposed to a strong northerly drift up the coast, which sweeps it northward and piles it up against the southern coast of Banks Peninsula. When the supply of shingle fails, the sea encroaches on the land, as, for instance, near Oamaru, where works are necessary for the protection of the town, and the neighbouring coast northward to the Waitaki River is being rapidly attacked. This river brings down an enormous amount of material, and in consequence the coast immediately north of its mouth is advancing in spite of a probably downward movement of the land, evidenced by the valleys in the dolerite plateau near Timaru, which have the outlets depressed below sea-level, and also by the presence of submerged forests on the coast near the mouth of the Pareora River. The gradual pushing-forward of the shore is evident right up to Timaru. The railwayline is placed for the greater part of its length on the level beach which has accumulated at the base of the cliffs, now removed some distance from the sea. Some idea of the importance of the drift along shore can be seen at Timaru itself, where several acres of land have been reclaimed by this agency on the weather side of the breakwater, which has checked the northerly movement of shingle at that point. This check can only be temporary, as no possible human works can obstruct for long the action of an ocean-current. Further north, towards the southern end of the Ninetymile Beach, the sea seems to be getting the best of the struggle, and the fringe of the old fan of the Rangitata and Ashburton Rivers is now being cut away. But further north still, when the Ashburton and, above all, the Rakaia Rivers have poured in their contributions of shingle, accumulation is rapid, as is proved by the presence of the great shingle-bank between Taumutu and Birdling's Flat, dividing the shallow waters of Lake Ellesmere from the deep sea. The shore-current has followed here a direct course across the mouth of a deep indentation of shallow water straight to the solid mass of Banks Peninsula. By it the northerly current is turned eastward into deeper water, and is thus rendered incapable of carrying its load of coarse material, so that this is piled by wave-action across the mouths of the bays facing south, which have now become the valleys known as Price's Valley and Little River. The fine detrital material is swept off into deep water, and some may find its way round the eastern side of the peninsula, and be dropped in the somewhat sheltered area on its northern side. The rate

of accumulation near Birdling's Flat must be extremely rapid, as old maps show that Lake Forsyth had in historical times a permanent outlet to the sea, and later still the seas used to break over the narrow spit of shingle which divided the lake from the sea; and more recently, within the memory of the present writer, this bank has increased in breadth by several chains. This increase has taken place in spite of a probably downward movement evidenced by the presence of submerged forests in Lake Ellesmere.

The absence of any similar spit on the south side of Banks Peninsula between it and the plains is strongly suggestive that they have not been beneath the sea since the present land-surface was formed, as spits would have formed there if it had been depressed below its present level. The presence of a strong northerly current closely hugging the coast, and preventing the accumulation of silts and similar non-permeable beds interstratified with the shingle, explains the general absence of artesian areas to the south of the Rakaia.

The volcanic mass of Banks Peninsula is a fundamental factor in the formation of the Christchurch artesian area, since under its shelter have accumulated the finer sediments interstratified with coarser material without which artesian conditions would have been impossible. Immediately to the north of it the width of the “flowing well” belt is the greatest, and it narrows down as the distance north and south from it is increased. No doubt the northerly extent would be greater were the hills of North Canterbury further away, for the supply of artesian waterfalls off as they are approached, and fails altogether in their vicinity. The general structure of the area, as subsequent examination will show, is that of irregular beds of shingle and sand parted by clay-beds. They are not arranged in basin shape in accordance with the usual text-book diagrams, but in beds with a general slope to the east.

The water finds its way into the gravel-beds and flows seawards, and in all probability it has an outlet beneath sea-level to the eastward of the present shore-line. When a pipe is put down to the water-bearing stratum the water follows the path of least resistance, and if the surface of the ground is below the effective head of water the well will be a flowing one. This seems to be the principle on which the flow is dependent.

It is probable that during the formation of the artesian beds there has been a struggle between two forces—viz., the gradual sinking of the land, as evidenced by the presence of peat-beds several hundreds of feet beneath the present area sea-level, and the gradual building-up of the plain by aggradation. In a limited area parallel to the coast-line marine beds may have been laid down, and as sedimentation proceeded the area has been covered up by shingle from the rivers.

Present Surface Conditions.

The former conditions can be most easily realized by taking the present features of the area. In the sea off the present coast-line are beds of sand and mud; in certain places, as the estuary of the Avon and Heathcote, patches of both occur in close conjunction. This estuary tends to silt up on account of the gradual increase in height of the half-tide flats by the action of Zostera and other halophytes entangling the finer sediment carried by the rising tide and incoming fresh-water streams. Thus over the sand-beds a coating of fine silt is deposited. Then again, along the coast and also in different places inland, probably marking old shore-lines, are rows of sanddunes. Between them are usually peat-bogs of no great depth, the coating

of peaty matter rarely exceeding 2 ft. in thickness. These bogs rest on beds of clay formed by the sediment which has been deposited by rivers and tides, but principally by the latter raising the sea-bottom to the hightide level. The area of bog and sandhill stretches back from the sea till the surface gravels of the plains are encountered, which have no regular boundary, but stretch out as tongues into the swamp and sandhill complex. An excellent example of this is to be seen at Addington, where a shinglebank reaches down to the eastward, passes into Sydenham, and crosses Colombo Street South near the Sandridge Hotel. This shingle-deposit marks the presence of an old stream of gravel, and others are found in many places reaching out more or less into the swampy country. This is the general type of land-surface which is found all over the area. In those parts close to the sea sandhills and peat-bogs predominate, but on going away from the sea the shingle becomes more and more important, till it forms the whole present surface of the country. The land in all probability possessed similar surface-features when the water-bearing beds were being laid down.

General Conditions of Artesians.

In the publications of the United States Geological Survey there are numerous excellent papers dealing with the scientific aspects of the flow of water in artesian wells. In one of the earliest of these, entitled “Requsite and Qualifying Conditions of Artesian Wells” (5th Annual Report, 1883–84), Professor T. C. Chamberlin considers the general conditions which determine the occurrence and the flow of wells. Other papers of great interest are those by F. H. King, on the “Principles and Conditions of the Movements of Ground-water,” and by Charles S. Slichter, on the “Theoretical Investigation of the Motion of Ground-waters,” both in the 19th Annual Report (1897–98). In a paper on “Rock Gas and Related Bitumens” (11th Annual Report, 1889–90), W. J. McGee expresses the conditions governing artesian wells so very succinctly that I quote his actual words, especially as they have an important bearing on the conditions governing our own wells. He says (page 603), “In the inverse order of their importance, the requisites for artesian flow of water are—(1) Conditions of structure, (2) conditions of texture, and (3) conditions of supply. The most favourable structural condition is the arrangement of the strata in the form of a basin, as in the Province of Artois in France, from which the flowing well takes its name. The rain-water falling on the rim of such a basin percolates through the strata to its centre, and there rises through natural or artificial openings to a height depending upon the difference in altitude between the area of supply (or catchment-area) and the head of the well; but the basin structure is not absolutely essential, and artesian flows are obtained from uniformly inclined strata (such as those constituting the Atlantic coastal plain) when the catchment-area is elevated considerably above the well-head, or when the strata either diminish in permeability in the direction of the inclination or extend far beyond the point of outlet —i.e., textural and other conditions may combine with structure to produce now where structure alone is unfavourable. The most favourable textural condition is found when the porous stratum extends from the catchmentarea to the part of the basin, trough, or monocline tapped by the drill, and is overlain (excepting in the catchment-area) by an impervious stratum. This condition, like the last, is not absolutely essential to artesian flow, since all rocks are more or less pervious, and if the difference between the

catchment-area and well-head is sufficient a slight flow may be obtained almost anywhere; but it is essential to abundant flow, and, indeed, to the slightest flow, unless other conditions are exceptionally favourable. The most favourable condition of supply is found when the porous formation rises from beneath less porous strata and forms the suiface over a broad arer.”

Aften this general statement of the conditions governing artesian wells we must pass on to consider those which affect the particular area under consideration.

Extent.

The artesian area referred to in this paper consists of the belt of country which fringes the coast from the mouth of the Ashley River southward through Kaiapoi, Christchurch, Tai Tapu, Ellesmere, to the mouth of the Rakaia. Its length is nearly fifty miles, but its breadth varies considerably, being very narrow at the north and south, and reaching its greatest width in the neighbourhood of Christchurch, where it is about ten miles wide. Its inland boundary is roughly marked by the railway-line which runs from Christchurch north to Leithfield and south to Southbridge. In one or two places the area extends slightly over this approximate boundary, the artesian well farthest from the sea being at Islington, on the Main South Line, about eight miles from Christchurch.

Structure of the Area.

and finer material, with occasional peat. There is a progressive diminution in the amount of gravel encountered in the bores on approaching the present coast-line. This is by no means uniformly true, as the Boys' High School well, in the western part of the city, shows a very large proportion of gravel to be present, a feature which is also shown to a minor degree by other wells in its vicinity—for example, the Christ's College, and the Exhibition well in Hagley Park. This exception does not, however, negative the statement that there is a general increase in the amount of shingle on going east. These beds must have been laid down where strong currents were in operation, due either to river or sea currents, the former in all probability; and they are almost certainly due to the aggrading action of a powerful stream on a land-surface. The presence of peat also proves that subaerial conditions obtained over the area while the beds were in process of being deposited. The sand and clay beds interstratified with them are in all probability principally of marine origin, since remains of shells are frequently encountered in them. Their association with land-beds proves that there was a struggle going on between the forces that tend to build up a sea-bottom to the level of the sea and to continue it as a land-surface above that level, and, on the other hand, a general sinking of the land, which is evidenced by the peat-beds now found so far below sea-level—600 ft. in the case of the well at Islington. It is apparent that at times the aggrading forces got the best of it. How far this depression of the land has gone on in excess of that already proved is quite uncertain, and only deeper wells will disclose the information.

The water-bearing beds of this section, as, indeed, is the case in others, are almost invariably composed of shingle. According to well-sinkers, water is frequently found all through these gravels, but the most prolific supply is obtained from just above the impervious layer below the beds; in fact, its general distribution throughout the gravel-layers in a particular well is looked on as an unpromising indication for a good flow being obtained from that bed. The overlying impervious layer is usually clay, but it may be sand or even more consolidated and less porous gravel.

The height to which the water rises is generally found to increase with the depth of the well, though in one or two cases the reverse is found in the case of a particular bed. This may be put down to friction preventing the passage of water through the bed. Owing to the level of the ground over a large part of the area being only a few feet above sea-level and sensibly uniform, there is apparently little difference in the height to which water rises on approaching the coast, but the condition of texture of the bed appears to be the controlling one affecting the height to which wells rise for a particular depth. In the part of the area to the west where the plains rise somewhat steeply the wells are non-flowing, although the pressure of the head maintains the water in the pipe at a fairly definite level, and it would flow, if it were possible to take it off, below the level of the ground. The water in the first stratum of the well at Islington rises only to 42 ft. of the surface of the ground—that is, the level of the water is 70 ft. above the sea.

Series No. 2.—Deep Wells: Central Ward, Christchurch.
(Plate X.)

This series of sections gives a record of a number of the deeper wells of the centre of the city, with one or two others added for purposes of comparison. They furnish the most detailed representation of any particular

locality available at present, and they suggest that when other areas are similarly plotted the minute structure of the area may be determined, and some of its apparent irregularities may be removed.

From this series of sections it will be seen that certain sands, sandy clay, and clay beds are persistent and regular over a fairly wide area. Well-marked beds of this type occur at a depth of between 40 ft. and 80 ft. below the surface, another between 150 ft. and 175 ft., another between 350 ft. and 400 ft. Similar, but less persistent, beds divide up some of the main gravel-beds into subordinate and irregular layers. From the occurrence of shells in certain of these beds it may be inferred that they are marine or aestuarine, although it is possible that some are not. It will also be seen that a bed of gravel, thick and continuous in one well, is frequently broken up in a neighbouring well by sand and clay beds. As nearly all these gravel-beds will yield artesian water, it is obvious that it is almost impossible to state the exact depth of the different water-bearing strata in any locality, as a particular water-bearing bed of one well may be divided up into a number of thinner beds in a neighbouring well, where, owing to different conditions of pressure, friction, and the supply of water, flows of varying amount may be obtained from what are apparently different beds, although they are really connected with one another. It is thus impossible to predict with any degree of certainty the number of water-bearing beds that will be encountered within a given depth, or that water will be met with at any particular level, though certain belts are extremely likely to contain it in one or more levels. It will be noticed that a very uniform and widely distributed water-stratum is found about the 400 ft. depth; all the wells from this show a marked regularity in the height to which the water rises above the surface. This is, in all probability, a stratum covered by marine or aestuarine bed, an occurrence which suggests that the conditions were uniform over a comparatively wide area.

This series of sections shows the presence of peat-beds, sometimes singly, and again divided by thin bands of clay and gravel. Remains of wood are also encountered.

The height to which the water rises in any particular well is also found to depend principally on the depth, although there are one or two departures from the rule, as mentioned in dealing with Series No. 1.

Series No. 3.—Papanui, St. Albans, Richmond, and Shirley.
(Plate XI.)

This shows the structure and character of the water-bearing beds in the district stretching from Fendalton, through St. Albans, towards the coast. The first well lies on the edge of the artesian belt, and shows, at all events in its upper levels, a marked predominance of gravel; but on going east the amount of gravel in the sections decreases, and finer detritus becomes more important. There is a certain element of regularity in the arrangement of the beds. A well-defined clay-bed extends almost all over the area, with an occasional coating of peat, and under it is a fairly regular band of gravel. Under this lies an extremely persistent bed of sand or sandy clay, which contains at times marine shells, thus showing that the sea stretched over the area at a comparatively recent date. This is succeeded below by a somewhat broken set of beds composed of gravel separated by peat and also by fine detrital matter; but in spite of the apparent irregularity there is an approach to order in the arrangement, and a more complete set of sections would show this absolutely. Even the sections as they

stand give a fair idea of the thinning-out and thickening of the same bed in different parts of the district. Below the belt where gravel predominates there is a set of beds composed of finer detrital matter, which would probably be continuous if the records of more deep wells were available. It will be noted also that there are very regular and widely distributed peat-beds, especially in the Fendalton and St. Albans area.

The series shows that there was a general struggle for existence between the land and the sea. The periods when the former got the best of it are indicated by the presence of peat, and perhaps by the gravel-beds, and the time when the sea stretched over the area is indicated by the finer-textured beds with their occasional shell-remains. The amount of rise in the wells increases, as a rule, on approaching the coast-line, and this may be put down to the better textural conditions near the coast. Instead of the water being distributed throughout the thick beds of shingle it is concentrated by the impervious beds into narrower bands, and so gives higher and stronger flows.

Series No. 4.—Sydenham, Opawa, Heathcote to Estuary.
(Plate XII.)

This series shows the structure of the belt of country which fringes the foot of the Port Hills and extends across the estuary towards New Brighton. The wells of special interest are those close to the hills, in whose records there is frequent mention of the presence of angular matter of volcanic origin. It is probable that the water supplying them comes from rain which falls on the Port Hills, and not from that on the plains, but more detailed work will have to be done before this statement can be maintained for certain. There is no doubt that the rocks of the hills exert a disturbing influence, since some wells which are sunk in their vicinity to beds of shingle yield no water, though a little distance further away similar beds at equal depths are prolific. These beds must be cut off in some way from the main artesian area. Others of them are down very close indeed to the underlying beds of volcanic rock, especially those near the present estuary, a fact emphasized by the record of the wells sunk by the Sumner Borough Council in their efforts to obtain water for the reservoir which supplies the borough. The ignorance of artesian conditions, as well as the unnecessary expenditure of the public money resulting therefrom, is thoroughly exemplified in connection with one well in the estuary near the Fisherman's Flat. After getting a poor flow of water at a depth of 416 ft., the boring was continued 41 ft. further, through layers of scoria and hard black basalt, evidently an outlying part of the rocks of the Lyttelton volcano, in expectation of getting a more plentiful supply. Needless to say, this hope was not realized.

The records of the wells near the estuary show that beds containing shells are met with very persistently, and it may be inferred therefrom that the conditions on the north side of that part of Banks Peninsula have not altered materially since the beds were first laid down—i.e., the area has been aestuarine for a very long time.

This series contains the record of two very deep wells: the first sunk to a depth of 708 ft., near the old Heathcote Racecourse, in order to obtain, if possible, a supply for the Lyttelton Waterworks—an unrealized expectation; and, secondly, the well sunk at the Sydenham Water-tower, which reached a depth of 572 ft., and gave a flow of water from a depth of 550 ft., which rose 32½ ft., and was described by the well-sinker (Mr. J. W. Horne) as the largest flow he ever got. I believe that this is the deepest flowing

well in the Christchurch artesian area, the well at Islington, though sunk deeper, not giving a flow at the surface.

Series No. 5.—Coastal Belt: Sefton to Estuary of Avon and Heathcote.
(Plate XIII.)

This series gives a selection of wells from along the coastal belt reaching from Sefton to the north of the Ashley River, through Kaiapoi and New Brighton to the Port Hills at the estuary of the Avon and Heathcote. In the extreme north of this belt the supply of water is poor, but fairly good flows are obtained in Woodend and at Kaiapoi; nevertheless, the supply is both deficient in amount and in the height of rise as compared with wells in the immediate neighbourhood of Christchurch. The relative small height of rise can be readily seen by comparing the Kaiapoi Borough Council well, sunk to a depth of 450 ft., with deep wells in the middle of Christchurch. The Kaiapoi well gives a rise of only 10 ft. from that depth, whereas the Christchurch wells give a rise approximating to, or even exceeding, 30 ft. This well gave a flow of 50 gallons per minute at a height of 1 ft. when first sunk, an amount which compares favourably with deep wells further south; but the comparative smallness of supply is borne out in the case of other wells. Inland from the town the supply falls off considerably, for at Ohoka a well sunk to a depth of 374 ft. gave only a 2 ft. rise. In this case, however, the beds were almost exclusively of shingle, and the conditions of texture were decidedly unfavourable for the production of much artesian water. As a general rule, the beds of Kaiapoi and its immediate vicinity show the presence of a large proportion of shingle, from which it may be inferred that a large river has occupied the present position of the Waimakariri for a considerable period of time. Just to the north of Kaiapoi, and also to the south, beds composed of finer detritus become more important, till when the estuary is reached they are almost wholly of fine material, with an entire absence of shingle.

The two important factors controlling the conditions of texture of the beds are the presence of Banks Peninsula to the south and the delta of the Waimakariri in the north. This river has pushed out its delta beyond the general trend of the coast-line, and the finer material brought down is deposited on either side of the mouth. This accounts for the area of swamp immediately to the north about Woodend, and also the extensive deposits of fine material in the neighbourhood of New Brighton. An important factor in the latter case is the strong littoral current which runs down the coast during northerly winds. At that time the Waimakariri is usually in flood and heavily charged with sediment, so that the conditions are eminently favourable for the transport south of large quantities of detritus. A part of this is carried landwards, and ultimately forms dunes along the shore; and another part is deposited in shallow water offshore, or in estuaries formed behind the sand-dunes. Such aestuarine deposits are indicated in the sections given by the wells by the frequent marine shells which are brought up from the bores. These show that on the site of the present estuary of the Avon and Heathcote the conditions have been the same for a long period of time. No peat or other land deposits are met with in this part of the district till deposits of angular matter are reached at the base of the series, derived without any doubt from the disintegration of masses of volcanic rocks on the Port Hills.

The wells along this belt of country close to the shore ise in sympathy with the tide, and the water contains a high percentage of salt.

Level of the Water in the Wells.

The height to which the water rose from the various levels in each well at the time of sinking is given in a small figure placed near the vertical sections at the depth at which the water was encountered. The surface of the ground is taken as the datum-line, and the height above this is reckoned as positive and below it as negative; the latter applies to non-flowing wells and to beds which do not yield a flow at the surface. These records furnish a basis for determining the amount of fall in levels which takes place after a well has been flowing for some time, a fall due partly to the fact that all wells go down slightly after a short period of time, even if they are not interfered with by others, and also to the fact that wells affect each other materially if they happen to be sunk so as to cause overlapping of the cones of depression which surround each well. The fall in level of the wells has been very rapid since they were first sunk in the district.

[The section below cannot be correctly rendered as it contains complex formatting. See the image of the page for a more accurate rendering.]

In his paper entitled “The Behaviour of Two Wells at the Canterbury Museum” (Trans. N.Z. Inst., vol. 28, p. 654, 1896) Captain Hutton states that a fall of from 2½ in. to 23/4 in. per year has taken place in the case of first-stratum wells, and as much as 5½ in. per year in the case of second-stratum wells. Comparing Captain Hutton's record of the height of the Museum well in 1895 with its height now, it is found that it has fallen off 53½ in. in the fifteen years which have elapsed, thus giving an average yearly drop of 3·6 in. It is therefore falling off at a slower rate now than when Captain Hutton made his observations. In the paper just cited Captain Hutton also gives his observations on the height of the wells relative to the weather-conditions, &c. He notes the rapid response to rain; an occasional lise, apparently inexplicable on any known cause; a marked evening rise, which he attributes to a lessening of the call on the wells towards evening. He failed to find any tidal effect or any trace of variation in sympathy with floods in the Waimakariri.

These observations are being continued at the present time by the author, and also by Dr. Hilgendorf, of Lincoln College, but it is too early to make any definite statement as to their trend, except that the influence of rain and the existence of the evening rise are undoubted. It was hoped that a variation in sympathy with the barometer might be detected, which might explain the evening rise to a certain extent, and also the non-response of floods in the Waimakariri, which usually come at a time of low barometer. It is possible that the effect of floods in the Waimakariri may be just sufficient to mask the effect due to low barometer, and the isolation of the two separate effects may be a matter of considerable difficulty. The present author has received from very reliable observers so many authentic statements of the influence of the Waimakariri on rams and wells that the idea cannot be set aside as quite without foundation, and further observations are necessary to establish or disprove it. The undoubted effect of the barometer on the flow of wells must also be taken into account, and future observations may settle the question. At the same time, it must be admitted that the observations at the Museum may lead to no very definite result, owing to the disturbing influence of neighbouring wells, but much more may come of those which are being made at Lincoln under the direction of Dr. Hilgendorf, since the wells there are comparatively isolated.

The greatest rise above the surface of the ground in the wells I have examined is that of a well in Manchester Street North, which reached as high as 37 ft. from a 451 ft. level; but a large number of others give a height nearly as much as this when sunk to about the same depth. As a general rule, the height increases with the depth, but there are a few cases where variations in permeability seem to exert a neutralizing effect on the depth.

Amount of Flow.

The amount of flow per minute is very variable, and lies between zero and 100 gallons, and it is possible that greater flows occur of which I have no record. As a general rule, the quantity is greater at the deeper levels of each well, though with marked exceptions. The well just cited above as giving a very high rise yielded 40 gallons per minute at the surface from a bed 451 ft. deep, a flow of 20 gallons from one 386 ft. deep, but one of 60 gallons from the 296 ft. level, thus showing clearly that the amount does not altogether depend on the depth. I am informed by Mr. Osborne

that such anomalies are frequent. The quantity falls off naturally if the well be tapped at a height above the ground, but in certain cases a very considerable flow is obtained even at a high level. Observations have not been made to see if the diminution of flow follows the law stated by Slichter in the paper cited previously, but it is hoped that they may be carried out.

It is evident that not only the depth but the permeability of the strata and the amount of supply affect the yield from any particular well. The well-sinkers say that the best flows are obtained a little above the lower boundary of the water-bearing stratum, and, further, that where the water is distributed throughout a water-bearing bed the flow is always poor. These facts tend in the direction of the idea that there are well-defined layers of water percolating, probably at a somewhat rapid rate, through the porous gravel-beds of the area. When one considers the number of wells, the large amount which flows from some of them, and the great waste of water that is going on continuously, it will be readily realized that they must derive the supply originally from some very prolific and constant source, and one which is not very materially diminished by the enormous tax on it.

Source of Water.

The water which supplies the wells appears to come from two sources—viz., the rainfall on the plains and the underground percolation through the shingle from such rivers as the Waimakariri. That the first source is undoubted is proved by the falling-off in the amount of water supplied by the wells during a period of dry weather, and their immediate recovery after rain. On one occasion the level of the water in the well at the Canterbury Museum jumped up 6 in after several days' rain. The same effect was noted by Captain Hutton. The response is so marked that there is no doubt of the source of the supply, nor that the permeable beds outcrop but a short distance—probably a few miles—from where the wells are sunk. However, as the rainfall over the Canterbury Plains in the neighbourhood of the artesian area is comparatively low, and has an average of about 25 in. per annum, with a minimum of 14 in., recorded to date, it is evident that this amount could not furnish a supply sufficient to satisfy the great drain on it and keep fairly uniform, were there not some more reliable and constant source. The only satisfactory explanation for this main supply is that there is a very large amount of percolation through the shingle of the plains from leaks in the river-beds, which finds its way across the out-crops of the permeable beds and thus contributes towards the artesian supply Such leaks do undoubtedly occur, and during dry seasons a very large proportion, if not all, of the water in certain of the shingle river-beds is flowing underground. At a small depth there is usually an abundant supply, even when they are perfectly dry on the surface, and it takes but a slight shower of rain to make such a river again flow on the surface. In some of the rivers, like the Selwyn, the water disappears completely from sight even during normal seasons, to reappear miles lower down at the surface where the underlying beds are not so permeable. Many springs also occur along a belt which runs approximately through the place where the Selwyn water comes to the surface paralle to the coast-line and near the upward limit of the known artesian area. It is probable that the containing impermeable beds of clay and hard sand reach their furthest

westerly limit about that line, though, by sinking deeper wells still, they may be found to reach much further west at greater depths.

It is absolutely certain that water sufficient for the supply flows through the shingle. A statement has been repeatedly made to me by reliable persons that their wells flow much better during and after a nor'-wester. Accepting these statements as true, though my observations on the Museum well do not confirm it, the explanation would be that during nor'-westers the rivers are high, owing to heavy rains and melting snows on the main ranges to the west, and so an additional supply would be expected in the underground streams formed by percolation from the river-beds, and therefore the wells would respond with an increased flow. If the observations on which this conclusion is founded are satisfactory, it would tend to confirm the idea that the major part of the water in the artesian area comes from underground streams which reach the edge of the outcrops of the clay-beds occurring on the eastern fringe of the plains in the neighbourhood of Banks Peninsula.

Tidal Wells.

Along the shore-line the height to which water rises is affected very markedly by the tide; for example, at New Brighton, according to observations carried out by the author, the level reached by water standing in an open pipe varied during a tide as much as 18 in. in the case of a first-stratum well (depth, about 144 ft.), while a second-stratum well (depth, 280 ft.) was only affected to the amount of 10 in., both wells being a few yards away from the high-water mark and slightly above it. The influence of the tide on the wells is noticed all along the coast from north of the mouth of the Waimakariri nearly to the mouth of the Rakaia, and its effect is felt inland for a distance of three miles in the neighbourhood of Christchurch, with a gradually diminishing amount as the distance from the sea increases. It is impossible to tell the exact limits of the influence of the tide, owing to local and variable causes masking it when the amount is small.

It is also noteworthy that these tidal wells are salt. Near the shore the water from the first-stratum wells is so salt as to be unpleasant to drink, while that from the second stratum is less salt, the amount of saltness falling off with greater distance from the sea. No data are at present available as to the amount of saline matter present in these wells, but this-will be furnished in a subsequent paper. In certain cases, too, wells sunk some distance from the sea are distinctly saline immediately after being sunk, but lose their salinity after an interval of a few weeks; but this does not apply to those near the shore, which are permanently salt.

Natural tidal wells are known from other parts of the world, and I have seen a reference to similar artesian wells round the shore of the Bay of Tokyo, in Japan. I cannot find the article on this subject which drew my attention, but, as far as I can remember, the writer attributed the variation in level in this case to the loading of the surface of the ground by an additional weight of water at high tide, and cited experiments as to the effect of artificial loading on the height of wells. As far as I can recollect, the two cases are somewhat parallel; but the presence of salt water in the Canterbury tidal wells suggests that these wells have access to the sea, which is not explained by the theory just mentioned.

The combined effect of the tide and the presence of salt may be explained if it is understood that the water-bearing stratum has an outcrop under

the sea, and that there is a possibility of mixture occurring between the salt water and fresh under these circumstances. It will generally be thought that an outlet under the sea would at once destroy the possibility of getting artesian water near the shore, but I think it can be shown that even with an outlet the presence of a counterpoise of sea-water may be almost as effective as an impermeable bed blocking the outflow. The conditions will resemble to a certain extent those of the well-known laboratory experiment of the U tube, with its branches filled with unequal lengthened columns of liquids of unequal density which balance one another. It is perfectly possible to draw off a supply of the lighter liquid from a level above that of the heavier liquid. If now we suppose that salt and fresh water are the two liquids, then if we have a constant supply of fresh water carefully introduced it will be possible to draw off a continuous supply of fresh water at a higher level. Let us now apply this experiment to the circumstances on the shore of the artesian area—say, at New Brighton. In this case the rise of the tide is about 6 ft., and the depth of the first-stratum wells near the shore about 144 ft. We may suppose, therefore, that the water-bearing stratum outcrops at an approximate depth of 144 ft. below low water. Taking the specific gravity of sea-water as 1·025, the length of a column of fresh water which would exactly balance this would be 144 × 1·025—that is, 147·6 ft. So that a well sunk at the level of low water and supplied with a continuous amount of fresh water from inland would flow at about 3 ft. above the surface. If now the tide rises, the well will rise with it. Theoretically, a rise in the tide of 6 ft. with no admixture of the liquids would cause the well to flow 6 ft. higher. The New Brighton wells do not vary as much as this, a discrepancy due to a certain extent to the friction which obstructs the rapid flow of water through the beds, thus diminishing the effective pressure, and also to the admixture of the salt water with the fresh at the bottom of the well, which also reduces the difference between the level of the sea and that of the wells. This mixture must take place owing to the disturbed conditions at the bottom of the well, due principally to the movement which must take place as the water is drawn off. This explanation accounts for the mixture of salt water with fresh in the case of wells near the shore, and also the falling-off in saltness as well as in the tidal effect on increasing the distance from the shore.

Conclusion.

It is hoped that the diagrams given with this paper may be of some interest and use to the general public, as they afford a certain amount of information as to the depths at which the water-bearing beds are to be found, and seeing that these records are put into such a form as to be readily understood.

The main results of this preliminary investigation of the area are to demonstrate,—

(1.) That the geological arrangement of the beds is not so irregular as was anticipated at first, but that certain probable marine beds are persistent over considerable areas.

(2.) That the water-bearing beds, being gravel, and in all probability laid down on a land-surface or by the agency of strong and varying currents, are liable to great variation in thickness and also to be split up by intercalated sandy or clay beds. This increases the number of water-bearing