The N.Z. Journal of Science and Technology, Vol. 30, Sec. B, No 30,  1948

By J.A. Bartrum and A.P. Mason, Auckland University College  

(Received for publication, 28th May, 1948)

Description is given of lapiez-like fluting[1] or gashes of solution pits which are wide, generally relatively short, vertical grooves with semi-circular cross-profile which end below in shallow basins.  Both types of phenomena are unusual in that they occur in basalt and not in limestone, which is their usual host.  The lapiez-like forms,  which have developed on broad steep faces of blocks of rock, are ascribed to down-flowing consequent streamlets fed by rain-drip from overhanging forest trees; dissolved organic acids have caused solution of the basalt along the courses of these streamlets.  The location of the miniature troughs so formed does not appear necessarily to be controlled by joints,  but should these latter more or less coincide in direction with the general slope of a rock surface, troughs which otherwise would be shallow may be accentuated into deep gashes.

Solution pits have a comparable origin, but in this case the drip has been temporarily localized at isolated centres and has fallen upon sub-horizontal rock faces. 

As early as 1916 one of the present writers (J.A.B.) noted outstanding examples of fluting, comparable with that characteristic of many exposed faces of limestone, in basalt in the valley of a stream which enters Hokianga Harbour about half a mile west of Horeke Township and which is followed by the Horeke Taheke Road.  He did not realize until the appearance of a paper by Palmer (1927) describing similar forms from Hawaiian basalts that such occurrences apparently were somewhat rare and therefore did not record his observations. This is now being done because of a recent opportunity of obtaining photographs to illustrate the phenomenon.

Lapiez-like Forms  
  The fluting that is illustrated by Figs. 1 and 2 is shown only on the moderately steep sides of fairly isolated, large blocks of basalt fallen on the two lower slopes of the stream valley in question, or onto its bed from a flow which has a maximum depth of not less than 60 ft. (19 m), although its full thickness has not been determined.  This flow overlies Upper Cretaceous sediments which in the vicinity of the main development of lapiez are likely to be soft shales, which are included amongst more resistant phases of the Upper Cretaceous sequence ;  actual outcrops, however, are absent from the particular locality.  By stream and subaerial erosion the sheet of basalt has been undermined and has shed blocks of rock, some over 30 ft. in average dimension, which a times constitute a chaotic mass at the stream.


The fluting is developed generally on notably steep faces of relatively isolated blocks of rocks which are seldom under 5 ft. (1.50 m) in diameter and which typically have a sharp ridge-like crest. In some blocks, as is shown by Fig. 1, it appears that the fluting and more particularly deeper gashes are located along joints which have served to permit differential weathering along their traces, although they have not been sufficiently open to allow disruption of the blocks in which they occur.  In other blocks, as can be seen from inspection of the photographs, the direction of fluting crosses joints obliquely and appears, therefore, to be independent of the latter.  The troughs of the lapiez are generally spaced about 5 in. to 6 in. (150 to 200 mm)  from axis to axis and are about 2 in. to 3 in. (50 to 75 mm) in average depth, unless excavated along joint fissures, when the depth may be much greater ; between the troughs there are sharp-crested miniature divides .


Fig. 1.  Lapiez in block of basalt near Horeke, Hokianga,  

Photo:  A.P. Mason

Fig. 2 – Lapiez as in Fig. 1  

Photo: J. A. Bartrum

Solution Pits

Wentworth (1944) has recorded solution pits from Hawaii which,  like those now described, typically are present on large, more or less isolated blocks of basalt or other igneous rock.  The majority of the solution pits at Hokianga resemble roughly halves of wide tubes which have been divided longitudinally (Fig. 3 and 4).  Possibly this extreme depth actually results from the union of members of a linked vertical series, for frequently these cavities occur one above another (Figs. 3 and 4) in a manner suggesting that a lower member has been formed at times by over-

Fig. 3 – Early stages of solution pits in block of basalt
near Horeke, Hokianga

Photo: J. A. Bartrum


flow of water from the basin next above it. In order abnormal example (Fig. 5) solution has led to a more or less disordered mass of hollows, some concave at the base but others irregular, separated one from another by irregular pillars or ridges with height usually not much in excess of 1 ft. (300 mm), although some have much greater relief.  In spite of the early impression of chaos in the pattern, there is a tendency to alignment of deeply etched hollows which are controlled in their location by joint which are evident in Fig. 5 to the left of and above the human figure.


Fig. 4 – Solution pit in block of basalt near Horeke, Hokianga

Photo: J. A. Bartrum

  Origin of Lapiez and Solution Pits

The conditions of occurrence make it impossible to ascribe the features described to other than chemical weathering as the primary cause of their production. They are unrelated to irregularity either of surface or constituent material, for the fairly fine-grained, generally ophitic and aphyric basalts in which they are developed are neither vesicular nor banded.

Yet the process responsible for the etching of the grooves and gashes must have been controlled by special conditions for the surface layers of the rocks concerned are fresh except where solution phenomena are lacking,  when some slight degree of chemical weathering at times is detectable.

    Bell and Fraser (1912, p.47) state that in rhyolitic fragmental rocks of the Waihi-Tairua region of Auckland Province “white walls of rock, frequently scored with vertical flutings or corrugations, recall familiar topography in limestone country.”

Fig. 5 – Solution hollows and intervening eminences in block of basalt near Horeke, Hokianga

Photo: J. A. Bartrum

  This brief statement suggests that this corrugation does not occur on fallen blocks ; in contrast, so far as the present writers have yet seen, in basalt it is limited to such blocks.  Besides the examples selected for description in this paper, fluted basalts occur fairly freely in valleys closely to these latter and have been reported[2] from other North Auckland localities and from the east end of Waiheke Island, Hauraki Gulf (E.J. Searle).  The fluted blocks invariably are large and have been shed by thick flows with major joints very widely spaced.  The size of such blocks would appear to be a pertinent factor in the development of solution pits phenomena, for, under the hypothesis of origin that is advanced, it would be necessary for the blocks to maintain a stable attitude for an unknown but certainly lengthy period which may well have exceeded a thousand years.  

It is probable that only blocks of unusually large size would normally attain stable positions after gravitating from higher to lower levels ; in addition it is believed that as a rule a certain minimum and not inconsiderable area of rock surface is demanded before the genetic process can come into operation.

Wentworth (1944) has suggested that drip of rain from trees has largely been responsible for the lapiez and other solution forms that he describes from basalts and other igneous rocks at Hawaii. 

Fig. 6.  – Exaggerated solution “pits ” in block of basalt near Horeke, Hokianga. The elongated furrow to the right of the hammer appears to consist of two conjoined “pits”.

Photo:  A.P. Mason

Such drip appears to be the explanation also of the comparable features of the Hokianga rocks,  for these latter invariably occur in areas where prior to deforestation there was a forest cover.  Drip from many New Zealand forest trees during heavy rain tends to be concentrated at some points rather than others,  falling with particular emphasis from such places as junctions of major branches and the undersides of large sub-horizontal limbs.  No rainfall figures are available for the Horeke district itself,  but rainfall maps (Kidson, 1937) show that on ranges close at hand,  which attain a maximum elevation of over 2’500 ft. (750 m), the mean annual precipitation may exceed 200 in. (5.10 m), although elsewhere in the Hokianga region it may fall as low as 60 in. (150 mm).  In North Auckland much of the rain comes during periods of heavy downpour which are separated by spells of a week or two during which the rainfall is slight.

During bursts of heavy rain the drip from trees can be particularly intense and will wash organic acids from the decaying bases of mosses, species of Astelia and other epiphytes which abound in the local forests. It is believed that these organic acids have a highly important role in decomposing the minerals of the basalts ; both olivine and lime-rich feldspar that is one of the main constituents of these rocks are likely to be decomposed readily by such acids.

Where the surface of a block of rock has been more or less horizontal,  drip concentrated into a somewhat cylindrical mass can be expected to dissolve initially a shallow cup-shaped hollow such as is shown in Fig. 6 on the right hand half of the rock figured. The hollow to the left in which the match box appears, is typical of the unusual solution pit ; its base is slightly concave so that it holds, during and following rain, a small, shallow sheet of water from which, in the present instance, overflow occurs at two points on the outer margin, the courses of the resulting streamlets being defined by dark linear patches in the photograph.  The writers suggest that the reason why a cylindrical hollow comparable I form with a stream pothole has not resulted, is that overflow from the lowest part of the rim of a growing cup-like hollow has continuously dissolved this part of the rim as deepening proceeded.  Such overflow may be responsible for the production of series of such forms in which one is perched above another,  but, as will be suggested later, another cause may operate.

In connection with typical lapiez shown by the basalts  it is true that in some instances weathering along joints almost certainly has early given rise to slight troughs which have then initiated small streamlets ;  these have passed down the rock faces and deepened such initial hollows largely by the chemical action of dissolved organic acids.  Inspection of Fig. 2 will show, however, that the direction of fluting may be independent of joints which ostensibly should have exercised directional control.  It is probable in this latter case that, on suitable inclined surfaces, rain drip has given rise to minute consequent streams which have recurred in the same embryo channels at successive bursts of rainfall and thereby have continuously deepened them into typical lapiez.  Hitchcock (1947) ascribes coarse fluting in granite in the Orinoco-Ventuari region of Venezuela to minute invisible joints, for in a few cases the fluting is parallel to joints along which the blocks have split apart.  Such joints as there are in the Hokianga rocks, however, are far from regular in their trend and do not appear to play an essential role.

The reason for the lack of appreciable weathering in the rock closely associated with the solution features described is apparently the materials disintegrated by the chemical activity of organic acids in waters passing over the rock surfaces are immediately carried off by the swiftly moving drip-fed streamlets which occupy the troughs and other hollows.

It has been suggested above that considerable time may be involved in the production of lapiez and solution pits.  It does not appear, however, to be essential that the drip from trees fall from identical spots during the whole of the process.  The individual trees of the forest must change and be replaced by others as time goes on.  Yet, once the formation of troughs or pits as well in hand should change of location of major rain drip be altered by change in forest conditions, much of the water would still find its way into the prepared depressions and extend them by solution.  Under suitable conditions, however, it seems likely that fresh solution pits could be developed on a flattish rock surface and it is highly probable that this is the explanation of some of the members of the vertical series of linked solution hollows that have been described.



Bell, J.M. and Fraser, C.  (1912) : The Geology of the Waihi-Tiarua Subdivision. 
NZ Geol. Surv.  Bull 15.

  Hitchcock, C. B. (1947) : The Orinoco-Ventuari Region, Venezuela
Geog. Rev., 37, 525-66

  Kidson, E. (1937) :  The Climate of New Zealand. 
Quart. J. Roy Meteorl. Soc., 63, 83-92

  Wentworth, C. K. (1944) : Potholes, Pits, and Pans
Subaerial and Marine.  J. Geol., 52, 117-30

[1] Some writers would use the term “lapiez“ for this type of feature only when shown by limestones.

[2] Not legible


How Kauris erode basalt

Dr. Neil E.Whitehead (ex GNS)

August, 2013

According to information you already have, the best guess from previous literature is that basalt at Wairere Boulders has erosion runnels in it because of the unusual presence of kauri. I have looked at the recent literature on this which confirms that, and there is a lot more known about this erosion now, although this is still not generally known even in the consciousness of scientists and must be collated from various sources.

First, it is a general property of Conifers that they promote podzolisation which is the turning of silicate rock into soil rich in silica i.e. erosion of silicate rock. One early author commented that kauri is probably the best agent for podzolisation of any tree, and creates an “eggcup” of this type of soil near the tree.   Fairly recent tests in the Waitakeres near Auckland did not show kauri could attack andesite well, (another common igneous rock) but this remains a very strong possibility for kauris on basalt.

Some detail of how this happens is now available. First laboratory studies directly on solubility of basalt have shown that it is increasingly soluble the more acidic the medium (liquid in lab tests, but in this case acidic soil). The soils under Kauri are exceptionally acid and this must promote erosion of basalt. The acid is partly tannic acids which are complex structure and high molecular weight.  (We see in a minute that the Kauri tannins are unusual. ) The acid is also partly small light organic acids, shown capable of combining with elements such as calcium in the rock.  Both these should act to slowly dissolve basalt. It takes 5000 years on average to produce viable soil from rock, and the rate might be expected to be higher for this type of attack. Thus the runnels  have been created in a few thousand years.

 Second, analysis of natural tannins in streams has shown that they carry many elements such as calcium which have come from breakdown of parent rocks. However the tannins produced by kauri and found in leaf litter are unusual, because they are combined with protein. This sequesters the nitrogen, stabilizes it and renders it accessible to young kauri seedlings or other plants which can grow close to Kauri, but inaccessible to other  trees which might compete with It, and also inaccessible to bacteria. The combination with protein has uncertain effects, but because nitrogen is present in the protein and exposed in many chemical configurations, it is quite likely that it may be unusually active in promoting dissolution of basalt by combining with (chelating) elements such as calcium and magnesium.   This would be like the action of an enzyme. This needs confirmation by experiment.

Any tannins originating from Kauri are more likely to be from the leaves, because the bark has been shown to have relatively little.

One mode of attack on silicates by Kauri has been shown to be stripping of a layer of silicate combined with aluminium. This leads to breakdown of the matrix.

Although literature search has revealed no parallels to Wairere Boulders, except the Hawaiian example already known, another potential place is some Pacific islands which are basaltic and also contain a close botanical relative of Kauri. A prime possibility is Vanuatu, but this does not appear to have been noted in the literature for further investigation.