Cadernos Lab. Xeolóxico de LaxeCoruña. 2008. Vol. 33, pp. 65 - 82

ISSN: 0213 - 4497

Description and nomenclature of the tafoni features (cavernous rock forms).

Research approaches in granite terrains

UÑA ALVAREZ, E. De 1

(1) Área de Geografía Física-Ourense (Universidad de Vigo). Pabellón 2-1º. Campus UniversitarioAs Lagoas. 32004-Ourense. E-mail: edeuna@uvigo.es

Recibido: 1/11/2007 Revisado: 8/4/2008 Aceptado: 20/7/2008

Abstract

Scientific terms have a descriptive, conceptual and categorical meaning. In the geomorphicresearch the applied terms can describe the landform features on the Earth’s surface; also, theycan reflect the different development conditions and stage of the studied landform; moreover,they are always related to the key concepts about the landform explanation. So, whole termi-nology has a very important role in the geomorphic research because supports the landformnomenclature. However, a suitable nomenclature according this finality must be unequivocal.But in the study of the cavernous features on granite rocks there is often a co-existence ofmany terms applied to the same form-types. Also, there is a literal or modified translation oflocal terms from regional studies. And the terms acceptance and their spreading by the resear-chers are very diverse along time. The situation can generate a conceptual and a categoricaluncertainty in the knowledge of the cavernous granite forms. This work deals with severalquestions about this terms meaning on the light to the research approaches in the graniteterrains. The earliest scenery where the landforms were recognised and termed dates from thelast years of the XIX century. Starting from this time, two general terms are usually applied toreport the basal or lateral cavities developed in granite rocks. The single word “tafone” (plu-ral “tafoni”) and the compound name of “cavernous weathering form”.

Key words: tafoni, cavernous landforms, terminology, typology, nomenclature, granite, geo-morphology

INTRODUCTION

The earliest news on basal or lateral hol-lows developed inside granite rock massesdates back a long time. The first descriptionwas collected by PRADO (1864) from Sierrade Guadarrama, in central Spain. A little laterthey were described from Corsica byREUSCH (1882): this author designed themwith the local term of tafoni (singular tafone)related to the verb “tafonare” (to make a holeor to open a window). They were reportednewly from Galicia, north-western Spain, byRagnar Hult (FRAGA et al. 1994) in 1898.Since then, these rock features are often des-ignated with the Corsican term. But theyequally are named with others local or partic-ular names in the scientific papers. The differ-ent morphological properties or the differentdevelopment stages from specific observa-tions over time are reflected in new terms.This terminology not only has a descriptive orcategorical sense but also involves a concep-tual framework selection in the study of gran-ite landforms.

In most cases the collective term has themeaning of an empty space in a rock volume,developed starting from a basal or a lateraldiscontinuity plane. The problem appearswhen it is necessary to summarize the knowl-edge available in the light of a new researchproject that needs the complete account of thislandform nomenclature. According to NOR-WICK & DEXTER (2002:11) “Long beforethere was a science of geomorphology peoplenotice the pitted weathering of some rocks...The nomenclature for pitted and cavernousweathering was not harmonized through mostof the twenty century but the word tafoni hasnow become standard for all such pits, largeand small.” It is noted the lack of agree withregard to the typical features that could definethis form. GOUDIE (2004) points out them asseveral cubic metres in volume, the arch-shaped entrances, the concave inner walls, theoverhanging margins (visors) and the smoothgently sloping debris-covered floors. But oth-

ers think that “the literature survey revealsthat there has been a great deal of liberty indeciding which features can be describedusing this term” (MIGON, 2006: 139).

The studies of granite terrains, wherethese landforms are well-developed and verycommon, had provided a differentiation intheir spatial, morphological and evolutionarypatterns (TWIDALE and VIDAL ROMANÍ,2005). Their names usually start from theirgeneral location in the granite landscape orfrom their local position in the host rock. Theyare also derived from their external visualappearance and from their inside features. Inaddition, they are related to the various originsthat have been proposed since they were rec-ognized in the granite terrains. All thesenames have been understood as the expressionof different landforms although this typologycan really reflect the different stages in thedevelopment and in the progressive degrada-tion of an initial cavity (VIDAL ROMANI etal., 2006). Moreover, this terminology hasbeen applied in many senses because theyhave been an unequal impact in the scientificcommunity. So, the progress towards a com-prehensive nomenclature system needs aresearch about their elaboration along thetime. This work is about this process to theforms developed on granite terrains. The useof the “cavity” and “hollow” words areapplied to an undetermined typological sensein this text.

FIRST DIFFERENTIATION OF THEORIGINAL TERMS

Tafoni, caverns, niches and cavernous rocksurfaces

The tafoni name was used or ignored in thepapers which made a field hollows descriptionuntil the twenty century three first set of ten.CHOFFAT (1895) applied it to mention thecases from the granite lands in the northern ofPortugal (Fig. 1a); this author wrote about themas a curious phenomenon of similar design with

CAD. LAB. XEOL. LAXE 33 (2008)66 Uña Álvarez

the forms named as tafoni by REUSCH (1882o.c.); also considered that these hollows wereequal to the first termed horados by PRADO(1864 o.c.) and identified as tafoni in non-gran-ite terrains by PENCK (1894).

“La colline de Faro d´Anha présente unoutre fait beaucoup plus curieux… cavitèsirrégulières, ayant à l´intérieur des bandesproèminentes, à arêtes vives, se trouvantaussi bien au toit que sur les côtés et parais-sant être les restes de parois ayant séparédes cavités distinctes. Ces ouvertures sontplus généralement sur le côté que dans lesdessus et se trouvent tantôt dans la roche enplace, tantôt dans des blocs isolés.” (CHOF-FAT, 1895 o.c.: 19)

All these cavities (tafoni or horados) wereexplained as the result of an exogenous processaccording to the granular disintegration or flak-ing observed from their inner walls. Alreadythere was a genetic concept in these terms. Andthis meaning became a favourable context tothe following use of two associated words intheir future designation: erosion and weather-ing. Soon after, CAYEUX (1911) describedmany cases developed in the gneiss and thegranite masses of Delos Island; their conceptu-al boundary related to the possible morpholog-ical types was lighted with their words:

“Entre le domaine continental, soumis al´influence exclusive de l´atmosphere, etl´étroite zone littorale façonnée parl´érosion marine seule s´étend parfois unaaire placée dans des conditions mixtes, oùles phénomenes d´érosion preséntent unintérêt de tout premier ordre...La nature y afaçonné les roches avec un art, une déli-catesse et une fantaisie uniques à ma conais-sance...L´érosion qui résulte de l´actioncombinée de la mer et de l´atmosphere estessentiellement une érosion alvéolaire, etparfois caverneuse...” (CAYEUX, 1911o.c.: 162)

The forms size was the key to provide aclassification (Table 1), excluding here theuse of the previous terms. But the explana-tions picked up in this work clearly retain asub-aerial origin. However, the author stated

that the two different types of cavities couldbe the result of two different evolutionarytrends. The smallest hollows with orderlypatterns named alvéoles were ruled over thegneiss surfaces near the sea: their fast growthmay merge them in a cavern. The largest iso-lated hollows named cavernes are usuallydeveloped on the inside of the inland graniteblocks (Fig. 1b): their growth starts from asingle cavity little by little differentiated(along tens of thousands of years). In spite ofthis, old morphological and evolutionaryconcept, later in most papers about thesegranite forms only appears a definition basedon the size.

CAD. LAB. XEOL. LAXE 33 (2008) Description and nomenclature 67

Fig. 1b Cavern cross-section from Cayeux (1911).

Fig. 1a Tafone cross-section from Choffat (1895).

Similar hollows founded in the semiaridand arid lands of the United States were relat-ed to weathering processes too. In this way,new terms were used to report the differentcavities. For instance, the shelters fromArizona and Nuevo México sandstone terrainswere named solution niches (BRYAN, 1928).In this context perhaps the most interestingcontribution to the following research in gran-ite terrains was the work of BLACK-WELDER (1929); he used to describe thevery frequent hollowing-out rocks fromCalifornia desert the general term of pockets,which has a meaning with regard not only totheir visual features but also to their develop-ment processes:

“Although commonly bowl-shaped orpurse-shaped, they display great variety ofform. One generally finds that the outsideof the partly decayed rock is more coherentthan the inner part...Excavation proceedslargely beneath such external crust. The

pockets are enlarged until they coalesceand thus leave isolated plates of the cruststanding out in relief...The cavities, whichhave been called “niches” by Bryan…arefamiliar features of most igneous rocks,sandstones, sandy shale and conglomer-ates.” (BLACKWELDER, 1929 o.c.: 393)

The pockets classification is equally basedon their size range (Table 1). But this authorsuggested that Bryan’s terminology onlycould be really used for the cavities developedby a single solution process. Against in a gran-ite cavity “the inner surface of the recess isgenerally covered with crumblingscales…identical with the usual products ofexfoliation… this condition favours theinward and upward growth of the cavity.”(Blackwelder 1929 o.c.: 395-396). This wasthe conceptual sense that reveals the term cav-ernous rock surfaces as a result of the cav-ernous weathering which has prevailed inUSA research until today.

CAD. LAB. XEOL. LAXE 33 (2008)68 Uña Álvarez

AUTHOR

Cayeux (1911)

Blackwelder (1929)

Bourcart (1930)

Centimetre range

Alvéolesen “dentelle”, “treillage”, “ruche”

et “éponge” dans les parois rocheuses

Cellules“cellules d´abeilles”

Pittedrock surfaces

Alvéoles“éponge de pierre”

Honeycombed rock surfaces

Metre range

Cavernes

Cavernous rock surfaces

Taffoni Grottes

Table 1. Main hollowing-out rock nomenclature at the beginning of XX century

Literal transcription and reinforced con-cepts

The term of tafoni was newly used in thedescription of granite cavities from Corsica.But BOURCART (1930) wrote the name as

taffoni (singular taffone), so retained long

time in the French research; just as this author

explain in the first page of their paper, the

word was transcribed literally from a publica-

tion about the Corsican landscape (Deprat,

1908). The features observed here were relat-ed in detail by the author:

“Les taffoni sont des cavités grossière-ment hémispheriques sculptées dans laparoi verticale des rochers de granite. Plusexactement, elles comportent une paroisupérieure, en voûte souvant très régulière,et un plancher quelquefois presque plan,mais toujours en pente vers léxterieur. Ceplancher peut être concave et la cavité prendalors une forme en demi-ellipsoide oumême en demi-sphere...Il arrive souventque deux de ces grottes ont empiété l´unesur l´autre. Au lieu d´être séparées par unpilier, comme c´est la cas quand elles sontsimplement juxtaposées, elles le sont alorspar un rebord tranchant. Deux taffoni peu-vent également êtres surperposés et le plan-cher qui les sépare peut s´effondrer; la cour-bure des parois de la nouvelle grotte qui enrésulte indique alors son mode de forma-tion.” (BOURCART, 1930 o.c.: 5-6)

Again the classification was size-based(Table 1). The smallest cavities were identi-fied as similar forms to the previous namedalveoli. The largest were recognised as niches.In fact, this work stated the same origin tothese types because both together were locat-ed on the same side wall rock surface. Sincethis reasoning the author noticed that an accu-rate genetic explanation of these forms isdoubtful: any initial condition could be erasedalong the evolutionary time, particularly in thebiggest cavities. So, the final statements ofthis study retain that these forms have anexogenous origin but they show a lack of con-sistency with the former research about therole of the rock structural properties in thegenetic and the evolutionary process. Eventhis time all studies underlined the signifi-cance of the joint system in the granite on thehollows shape whereas this author reveals thatthe control could be only possible in somespecific observations.

The following research was centred on thelandform classification related to their causalmechanisms. Some prior concepts and theirassociated terms were ignored. Another con-

cepts and names were reinforced. The differ-ent impact of these works on scientific com-munity gave rise to paradoxical interpreta-tions. For example, the hollows found in thegranite blocks from Monte Alvarado inPortugal north (clearly tafoni from severalphotographs) were named blocos cavadosbecause “os tafoni sao tipicamente litorais”(COTELO, 1940: 3) and the location of theseforms was far away the sea.

TERMS DIVERSIFICATION RELATEDTO MORPHOLOGICAL TYPES

The basal, the sheet and the side-wall forms

At the same time that this landform occur-rence began to be considered as a commonsuccess in the granite landscapes, the specificterminology applied to a suitable definition oftheir typology increased. In a general sense,all the new terms contained the concept of aprocess-response system with a dynamicalbehaviour under open air conditions. Theyhave an obvious descriptive role related to thecavity location in the unities of a granite mor-phological system (boulder, block, sheet,walls...). Also, they have a categoric role see-ing that indicate their size variations and theirgrowth stage. The research carried out in thegranite terrains of Corsica where “les taffonis´alignent systématiquement sur les princi-pales directions des diaclases” (LIGUS, 1952:185) and “every boulder of stone and also theoutcropping rock is perforated with cavities”(WILHELMY, 1958: 155). These featureswere object of a comparative analysis with thehoneycomb forms founded on any crystallinerocks from France (CAILLEUX, 1953) andfrom the northeast of Spain where “le niveaude base des taffoni est une surface de discon-tinuité, plus ou moins horizontale”(DENAEYER, 1953: 205). But also they werestudied in other locations as Brazil where theywere related to the structural properties of thebornhardts (Fig. 2a): here “les taffoni anciensqui s´étaint normalement dèveloppés à la base

CAD. LAB. XEOL. LAXE 33 (2008) Description and nomenclature 69

sont maintenant renversés...ayant son ventreen l´air” (BIROT, 1958: 32). The hollows areoften “roches évidées de façcon à offrir auregard une face intacte et une face creuséed´une grotte don’t l´intérieur est plus haut quel´orifice d´entrée, de sorte qu´elle est protégéepar une sorte d´auvent rocheux” (RONDEAU,1958: 453) (Fig. 2b).

The most successful nomenclature, stillmanaged today, was provided by severalregional studies: the hollows types were des-ignated according to the position of an initialdiscontinuity plane of growth in their hostunity (PANZER, 1954; KLAER, 1956; WIL-HELMY, 1958 o.c.; RONDEAU, 1961). Theimpressive use of this proposal is derivedfrom their apparent simplicity to the fieldwork. The basic terms that differentiated thehollows types were the basal, the side-walland the sheet forms (Table 2) but could bemore detailed. In order to define the shapesub-types, it was measured the hollows open-ing diameters and their perpendicular axis(Fig. 3). All these properties represent anobservational tool about the general “forms ofcavernous weathering, chiefly found in medi-um and coarse grained, acid to intermediatecrystalline rocks” (JENNINGS, 1968: 1103).

CAD. LAB. XEOL. LAXE 33 (2008)70 Uña Álvarez

Fig. 2a Cavity cross-section from Birot (1958).

Fig. 3 Depth measures in basal (A) and side-wall (B)forms.

Fig. 2b Cavity cross-section from Rondeau (1958).

In other granite regional studies thenomenclature was also based on the tradition-al size rule. The cavities from the desert landsin Chile were classified by SEGERSTROM &HENRÍQUEZ (1964) in two main groups: the

small centimetre forms or pits with a waffle-like patterns and the large metrics forms orcaves with diverse morphological features;these latter were usually perforated with owl’seyes forms. DEMEK (1964) distinguished inthe Bohemian massif between the largest side-

wall forms as the rock niches (wider thandeep) or the rock hollows (deeper than wide)and the smallest side-wall/basal forms as thedew-holes. Also, this author observed in theirfieldwork some small pits patterns namingthem with the old term of honeycombs.

CAD. LAB. XEOL. LAXE 33 (2008) Description and nomenclature 71

LOCATION

Beneathboulders or blocks

Betweensheet or bed planes

Insidesloped rock walls

(largest forms)

On whateversloping bedrock(smallest forms)

LANGUAGE

GermanFrenchEnglish

FrenchEnglish

GermanFrenchEnglish

GermanFrenchEnglish

TERMINOLOGY (SINGLE FORMS)

Basistafone (Hohlkehlen)Taffone de boule

Basal tafone, Basal niche, Basal cavern

Taffone de diaclase (horizontal)Sheet tafoni, Basal cavern

SeitentafoneTaffone de parois, Taffone de diaclase

Side tafone, Sidewall tafoneRock niche, Rock hollow

NebentafoneAlvéole, Nid d´AbeilleHoneycombed forms

*from Panzer (1953); Klaer (1956); Wilhelmy (1958); Rondeau (1961)

Table 2. Main cavities types according to the development plane into host rock*

Later research from granite rocks in theChile desert (Fig. 4) equally retained a sizedistinction from the honeycombs to thetafoni although “on ne peut douter qu´ils´agisse de la meme forme plus ou moinsévoluées” (GRENIER, 1968: 195). The pre-vailing concept was always an exogenousorigin: the basal type beneath sheets, alongjoints planes where moisture is retained, orto the side-wall type on vertical rock facesassociated with the ground level weathering,provided the same explanation to the cavi-ties found in the granite of Australia(DRAGOVICH, 1969), both clearly differ-ent of the small hollows groups named alveoli. Fig. 4 Cross-section of cavities in oldest development

stages from Grenier (1968).

At the beginning of 1970, really the avail-able observations and the data of these land-forms in the granite terrains were very numer-ous. They were noticed from all geographicalenvironments (i.e. CALKING & CAILLEUX,1962; WILHELMY, 1964; TSCHANG, 1966;PREBBLE, 1967). So, there was a need oflimit and define the research landform andtheir operational nomenclature.

Looking at a new suitable typology

The landform terminology now has a strictsense as a specific weathering landform in thegranite landscape. This meaning closes theresearch to many causal processes. On the onehand, this concept involves the two mainclasses of hollows typology derived by thesize and the patterns which were used untilthis time. A differentiation of prior namedalveoli or honeycombed surfaces (Photo 1)was made with more or less detail, not onlyabout the general terms but also on the partic-ular origins, for instance:

“Cavernous weathering may be used asa general term for the many landforms madeby weathering on steep slopes...occurs inmany kinds of rock, in many climates an onmany scales. Some cavernous weatheringpits are know as ´tafoni´...The smallestkinds of cavernous weathering need a dif-ferent term, and ´honeycomb weathering´seems popular...Honeycomb weatheringdevelops rapidly, perhaps in just a fewyears.” (OLLIER, 1969: 236-237)

“The nomenclature has not been stan-dardised and authors have variouslydescribed this type of erosion as ´alveolarweathering´, ´stone laticce´ and ´stone lace´Alveolar weathering has been preferredamong French geomorphologist while pub-lications in English often describe these fea-tures as honeycomb weathering or fretting.Commonly weathering features of this typeare described as miniature tafoni althoughthis usage is unfortunate since it implies thathoneycomb is genetically related to largescale cavernous features (tafoni). Although

tafoni and honeycomb weathering frequent-ly occur together numerous locations existwhere each occur independently suggestingpossible differences in their mode of origin”(MUSTOE, 1982: 108)

On the other hand, many evidences on thecomplex features and the frequent compoundhollows provided other morphological terms.The cavities were recognised not only underthe side of boulders or into the walls bedrockbut also into the curvilinear and laminar struc-tures, being very frequent on the steepenedslopes or flares where they were regarded asstages in a sequence of development from theweathering front evolution (TWIDALE,1971). The frequent occurrence of ruined orinverted forms endorsed that a hollowing outrock is a natural irreversible process. In thisway, the previous nomenclature position-based was multiplied according to the typesenclosed by the field measures. Incidentally,this shape was explained outside the concep-tual model of an origin only by weathering.For instance, the review of TSCHANG (1974:43) pointed to the relationships between theinter-feet distance and the curvature of thebasal forms (Fig. 5): the shorter distance isrelated to the higher curvature.

CAD. LAB. XEOL. LAXE 33 (2008)72 Uña Álvarez

Fig. 5 Support distance and cavity curvature in Hong-Kong from Tschang (1974).

At this scenery the book about the granitelandforms wrote by GODARD (1977) sum-marizes the morphology of these forms withthe following words:

“Creusés dans les parois rocheuses ouévidés aux dépends des flancs des grandesboules, les taffonis sont des cavités

arrondies dont la taille peut aller de celled´un oeuf à celle d´une grande chambre. Onen connaît des géants qui mesurent près de10 m de hauteur et occupent un volume deplusiers dizaines de mètres cubes...mais lesdimensions les plus fréquentes sont d´ordremétrique.” (GODARD, 1977 o.c.: 112)

Considering the size range of landforms inthe granite landscapes, this author identifiedthe cavities with the minor rank level: the sizeis a criterion both to discriminate and toexplain the landforms. These minor forms ormicromodelés from a granite outcrop allowsunderstanding the geomorphologic evolutionbecause many generative events could be dif-fered between them. Afterwards, when thisline of argument was retained there ran intothe question of the landform convergencewhenever an explanation of genetic processescame next to an exogenous natural system.

The following monograph work about thegranite landforms (TWIDALE, 1982) recog-nised the previous hollows dimensional rangeor types and also reflected the theoretical andpractical unambiguously of their nomencla-ture. And many studies began to planning adifferent perspective in the conceptual andmethodological sense of the accurate terms.

The research in the granite lands from Galicia(northwest of Iberian massif, Spain) made atthis time an interesting contribution set(VIDAL ROMANÍ et al., 1979; VIDALROMANÍ, 1983; VIDAL ROMANÍ, 1989).The hollow forms were named with theGalician term of cacholas or cacheiras. Inthese works the applied terminology attendedto the main conditions that could affect theshape and the size, the growth trends or theeventual destruction of whatever rock cavity(Table 3). First, the main control variable onthe dimensional and morphological features isrelated to the available rock volume that couldbe disintegrated inside the host unity which islimited by a discontinuity system; this controlon the size and the shape could be a heritagefrom an endogenous system. Second, the hol-low begins their development starting fromthis initial pattern on a single inner rock sur-face inside the host rock volume; since then,the main control variable concerns to the feed-back relations between the structural and envi-ronmental conditions: a diverse differentiationusually appears in an active and non-activeinner rock surface with regard to the weather-ing progress; later events can destroy the ini-tial hollows pattern.

CAD. LAB. XEOL. LAXE 33 (2008) Description and nomenclature 73

TERM

cachola pena

cachola laxecachola lapa

cachola fungo

MAIN FEATURES

Equal-dimensional host blockWell-defined rounded cavity

Common inner alveoliThe largest forms

Hetero-dimensional host blockHollow enlarged between joints

Basal or lateral locationOccasional inner alveoli

Steeped or flared host rockAsymmetric hollowing out wall

Rarely inner alveoli

EQUIVALENT TERM

Basal tafoneBoulder tafone

Tortoiseshell rocks

Basal tafoneSheet tafone

Side-wall tafone

Scarp foot caveMushroom forms

*from Vidal Romaní et al. (1979); Twidale (1982); Vidal Romaní (1983 & 1989)

Table 3. Hollow types related to structural control (Galicia, Spain)*

The successive stages of the hollowingout in a host rock had been really namedwith different terms. This assertion may beproved in the research at this time, forinstance from Italy (MARTINI, 1978) andfrom Finland (KEJONEN et al., 1988).

However, in the last monographic publica-tion of the twenty century dealing to thegranite landforms (VIDAL ROMANÍ andTWIDALE, 1998) the main form types aredescribed to look at a new conceptual senseof their nomenclature.

CAD. LAB. XEOL. LAXE 33 (2008)74 Uña Álvarez

Photo 1. Alveoli pattern in the OnsIsland (Galicia, Spain).

Photo 2. Complex of cavities in Achiras granite (Argentina).

TERMINOLOGY RELATED TO GRAN -ITE FORMS LINEAGE

The mentioned terminology was under-stood as a descriptive and a conceptual toolaccording to the knowledge progress about thegranite landforms. Generally, it was assumedthat the different stages of the recorded formswere single components of the whole land-scape. From this reasoning the cavernous

rocks are identified as limited forms to bestudied. The cited reports displayed their par-ticular conditions, selecting the shape and themeasure range starting from the most unusualcases. So, it was very difficult to explain thegenetic and the evolutionary events related tothese forms. However, if we consider theseforms as the components of a geomorpholog-ic continuum, their measures can provide asignificant information (Fig. 6).

CAD. LAB. XEOL. LAXE 33 (2008) Description and nomenclature 75

Fig. 6 Evolution of a granite slope with cacholas from Vidal Romaní et al (1984).

Found at many positions and climatic con-texts, the development pattern of the cav-ernous features involves a complex progres-sion from fresh to failed granite volume(Photo 2) reflecting cumulative changes in thestructural and in the mineralogical rock prop-erties. Today, the research clearly works on aself-destructive dynamical model of the hol-lows growth. This progression really proceedsin time periods where different evolutionaryphases are more or less fit to a sustainablegrowth in size (UÑA ÁLVAREZ, 2004). Theduration of each of these phases can be veryvariable and the current morphology of cavi-ties in the field indicates a partial self-organi-

zation system, also subject to divergent path-ways. The essential nature of their evolution isalways an irreversible process (Fig. 7).

The terminology meaning could be lessuncertain if the cavities are considered as mul-tistage entities from a holistic model, the nat-ural system. The key provided by the spatiallocation can be a valuable criterion fromminor to medium topographic and morpho-logic scale; the key provided by the develop-ment stage can be a valuable criterion to along time scale. But the derived nomenclaturedoes not design unequivocally the wide hol-lows types observed according to the size, theinner walls stage, the preserved visor features

or the shape of the opening in the outer shellof rock (Photos 3, 4, 5). In order to define anomenclature system associated with this con-ceptual sense, it is necessary to open thegenetic and evolutionary model prevailinguntil today. It is necessary to think about thecavities morphological features as the devel-opment stages associated to many impliedevents with a granite outcrop history.

Therefore, we find in the field research agreat variety in their location and their shapeproperties. For instance, a wide diversityappears in the features of their inner wallswhich can be regular, flaked, ribbed, honey-combed or scalloped, so that we must consid-er the possible generative fields with regard toa possible lineage of granite forms (VIDALROMANÍ et al., 2004).

from latent to primary hollows forms, latersubject of weathering modifications (VIDALROMANI et al, 2006 o.c.). The vast majorityof hollows inverse or laterally disposed ongranite rocks clearly start from a fracture planewhatever their orientation and some few on

CAD. LAB. XEOL. LAXE 33 (2008)76 Uña Álvarez

Fig. 7 Development stages in a boulder tafone (fromVidal Romani & Yepes, 2004).

The shape, the size and the different devel-opment stages of the cavities may be related tothe behaviour of the main control variablesalready mentioned (Fig. 8). In fact, only so canbe recognised the range of generative condi-tions and their features related to the age(s)

Fig. 8 Relation between block size and cavity area in Galicia (Spain).

weathered surfaces or xenoliths (TWIDALEand VIDAL ROMANÍ, 2005 o.c.). This condi-tion in most cases is field present: the cavitiesare related to the joint system which limitstheir host rock volume; and a recent empiricalresearch demonstrates that the size and thegeometry of this host rock have a significantcontrol on the size and the shape of these cav-ities (UÑA ÁLVAREZ and VIDAL ROMANÍ,2006). The generic approach supposes thatseveral morphogenetic fields can explain thecontrol conditions related with the origin andthe development of the hollow forms. So, theforms are termed according to the fieldresearch at an endogenous, subsoil or surfaceenvironment. The printed tafoni or latenttafoni name the lacunars spaces delimited by

the rock discontinuities at the apical or innerzone of a granite outcrop. Later evolution ofthis initial pattern can be recognised in the pri-mary endogenous tafoni at the granite land-scape. The primary exogenous tafoni can alsoexist associated to the natural relationshipsbetween a jointed rock volume and the subsoilor the external agencies. Whatever of theseforms can be degraded in secondary tafoniforms until their last ruined. Along this entireevolutionary path only some cavernous formsretain their generic design, for instance a singlecavity basal, reflecting an invariant morpho-genetic set. In order to identify the possiblemorphogenetic type, we need to go on with amixed spatial, morphological and statistical-dimensional research approach.

CAD. LAB. XEOL. LAXE 33 (2008) Description and nomenclature 77

Photo 3. Cavernous form very developed in Achiras granite (Argentina).

CAD. LAB. XEOL. LAXE 33 (2008)78 Uña Álvarez

Photo 4. Complex cavernous form in Achiras granite (Argentina).

Photo 5. Basal form with lateral window in Baroña, Galicia (Spain).

FINAL STATEMENT

The tafone or cavernous granite forms typ-ically show a size range from several centime-tres to several metres in opening diameter andtheir perpendicular axis. Such variation of thesize can provide a confused genetic meaningrelated to the identification of some alveolarforms. The alveoli are really discriminating by

their small dimension but over all the morpho-logical key concerns to their inherent closedpattern (honeycombed). And in the conceptu-al sense of the terminology we are usuallydealing to inner alveoli of the cavities whichreflect an anisotropic development stage.These forms were already explained with adifferent morphogenetic approach at the twen-ty century beginning.

CAD. LAB. XEOL. LAXE 33 (2008) Description and nomenclature 79

Tafone (Tafoni)Taffone (Taffoni)

Basal tafone (or taffone)Boulder tafone

Sheet tafone (or taffone)Sidewall tafone (or taffone)

Diverse opening plane shapeCross-section in arch shapeInner walls concave surfaceDiverse external walls stageConnecting to the outside

Cavern (Cavernous weathering)Cavernous rock surface

Basal cavern, Basal caveTortoiseshell rock

Rock niche, Rock hollowScarp foot cave, Mushroom rock

spherical, hemispherical, elliptical, irregularhemispherical, hemi-ellipsoidal

regular, flaked, honeycombed, ribbed, scallopedfluting, hollowing, preservation of visor or hood occasional by windows in the outer shell of rocks

Table 4. The most common terms and features of granite cavities studied.

USEFUL TERMINOLOGY APPLIED

MAIN FEATURES FIELD OBSERVED

The description and the nomenclature ofthe cavities studied in granite terrains overtime notes many common features from themorphological types differentially named(Table 4). At successive research approachesthese features were included in a nomencla-ture scheme based on several key criterions(i.e. location in host rock, development stage).Also, they were used to explain the cavitiesfrom a single or mixed action of the exoge-nous agencies. A similar approach wasretained to these landforms research in non-granite terrains where the morphological fea-

tures closed by the terms are also confused;therefore, similar terminology problems arepresent in non-granite terrains. And in thiscase the researches equally noted that in thehollows pattern there is a control “by the inter-nal fabric of the rock on where tafoni devel-op” (McBRIDE & PICARD, 2000: 878). So,a different conceptual and a different method-ological approach are necessary. Some formerideas and some current prospects mentionedcan provide the keys in order to clarify thefuture nomenclature framework.

DRAGOVICH, D. (1969). The origin of cav-ernous surfaces (tafoni) in granitic rocksof southern South Australia. Zeits.Geomorph., 13: 163-181.

FRAGA VÁZQUEZ, X.A., RIKKINEN, K.,VIDAL ROMANÍ, J.R. (1994). The visitof the geographer and botanist RagnarHult to Galiza. Ingenium, 4: 1-67.

GODARD, P. (1977). Pays et paysages dugranite. Paris, P.U.F.

GRENIER, M.P. (1968). Observations sur lestaffonis du désert chilien. Bull. Assoc.Geogr. Fr., 364/365: 193-211.

GOUDIE, A. (2004). Tafoni. In Encyclopediaof Geomorphology, London, Routledge,vol. 2: 1034-1035.

JENNINGS, J.N. (1968). Taffoni. InEncyclopedia of Geomorphology, ReinoldBook Co., p. 1103.

KEJONEN, A., KIELOSTO, S., LAHTI, S.I.(1988). Cavernous weathering forms inFinland. Geogr. Ann., 70A (4): 315-322.

KLAER, W. (1956). Verwitterungsformen imgranit auf Korsika. Geogr. Z., 33: 247-260

L.I.G.U.S. (1952). Problèmes géomorpholo-giques corses. Rev. Géom. Dyn., III (4):157-198.

McBRIDE, E.F. and PICARD, M.D. (2000).Origin and development of tafoni inTunnel Spring Tuff, Crystal Peak, Utah-U.S.A. Earth Surf. Proc. Land. 125: 869-879.

MARTINI, I.P. (1978). Tafoni weatheringwith examples from Tuscany, Italy. Zeits.Geomorph., 22 (1): 44-67.

MIGON, P. (2006). Granite landscapes of theworld. Oxford University Press, 384 p.

MUSTOE, G.E. (1982): The origin of honey-comb weathering. Geol. Soc. Am. Bull.,93: 108-115.

NORWICK, S.A. and DEXTER, L.R. (2002).Rates of development of tafoni in theMoenkopi and Kaibab formations inMeteor crater and on the Colorado

CAD. LAB. XEOL. LAXE 33 (2008)80 Uña Álvarez

REFERENCES

BIROT, P. (1958). Les dômes cristallines.Paris, Mém. et Doc. CNRS, Tome VI: 9-39.

BLACKWELDER, E. (1929) Cavernous rocksurfaces of the desert. Am. Jour. Sc., XVII(101): 393-399.

BOURCART, J. (1930). Le problème des taf-foni de Corse et l´érosion alvéolaire. Rev.Géogr. Phys. Géol. Dyn., III (1): 5-15.

BRYAN, K. (1928). Niches and other cavitiesin sandstone at Chaco Canyon. Zeits.Geomorph., 3: 125-140.

CAILLEUX, A. (1953). Taffonis et érosionalvéolaire. Cah. Géol. Thoiry, 16/17: 130-133.

CALKING, P. and CAILLEUX, A. (1962). Aquantitative study of cavernous weather-ing (taffonis) and its application to glacialchronology in Victoria Valley, Antarctica.Zeits. Geomorph., 6: 317-324.

CAYEUX, L. (1911). Description physique del´Ille de Délos. In Exploration archéologi-que de Délos - École française d´Athènes,Paris, Fontemoing, I (4): 162-186.

COTELO, J.M. (1940). Alguns aspectos ero-sivos dos granitos do norte de Portugal.Pub. Lab. Min. Geol. Porto, XIV: 3-8.

CHOFFAT, P. (1895). Notes sur l´érosion enPortugal. Sur quelques cas d´érosionatmospherique dans les granites du Minho(Tafoni). Com. Serv. Geol. Portugal, III:17-26.

DEMEK, J. (1964). Slope development ingranite areas of Bohemian massif(Czechoslovakia). Zeits. Geomorph., 8:82-106.

DENAEYER, M.E. (1953). Erosion alvéolai-re et taffoni du Cap de Creus (Province deGerona-Pyrennées orientales catalanes) etautres lieux. Bull. Soc. Belge Geol., 62 (2-3): 194-210.

plateau, Northeastern Arizona. Earth Surf.Process. Land., 27: 11-26.

OLLIER, C. (1984). Weathering. London,Longman, 2ª ed. (first in 1969, Oliver &Boyd)

PANZER, W. (1954). Verwitterungs undabtragungsformen im granit von HongKong. Mortesen-Festschrift, Veröff.Akad.: 41-60.

PENCK, A. (1894). Morphologie derErdoberfläche. Stuttgart, Engelhorns, VolI, 214 p.

PRADO, C. de (1864). Descripción física ygeológica de la provincia de Madrid. Col.Ciencias, Humanidades e Ingeniería, 2:60-76.

PREBBLE, M.M. (1967). Cavernous weath-ering in the Taylor Dry Valley, VictoriaLand, Antarctica. Nature, 216: 1194-1195.

REUSCH, H. (1882). Notes sur la géologie dela Corse. Bull. Soc. Géol. France, XI: 53-67

RONDEAU, A. (1958). Corse et Bretagne: àpropos du modelé granitique de la régionde Trégastel (Côtes-du-Nord). Norois, 5(20): 451-457.

RONDEAU, A. (1961). Recherches géo-morphologiques en Corse. Paris, ArmandColin

SEGERSTROM, K. and HENRÍQUEZ, H.(1964). Cavities or tafoni in rock faces ofthe Atacama desert, Chile. U.S. Geol.Surv. Prof. Paper 501-C: 121-125.

TSCHANG, H. (1966). Geomorphologicalobservations on weathering forms in HongKong and some other humid regions of SEAsia. Chung Chi Journal, 5 (2): 206-226.

TSCHANG, H. (1974). Geomorphologicalobservations on the tafoni forms of HongKong. Chung Chi Journal, 13 (1): 32-51.

TWIDALE, C.R. (1971). Structural land-forms. Camberra, Nat. Univ. Press, 247 p.

TWIDALE, C.R. (1982). Granite landforms.Camberra, Wiley & Sons, 372 p.

TWIDALE, C.R. and VIDAL ROMANÍ, J.R.(2005). Landforms and geology of graniteterrains. Balkema, Amsterdam, TheNetherlands, 352 p.

UÑA ÁLVAREZ, E. DE (2004). Tafoni enrocas graníticas. Primera valoración esta-dística sobre tasas de desarrollo en elMacizo de Ourense (Galicia, NW de laPenínsula Ibérica). Cad. Lab. Xeol. Laxe,29: 265-289.

UÑA ALVAREZ, E. DE and VIDALROMANÍ, J.R. (2006). Estructura yforma. Estudio experimental en cacholas(tafoni) de Galicia (NW peninsular). IXReunión Nacional Geomorfología,Santiago de Compostela, Actas: 785-793.

VIDAL ROMANÍ, J.R. (1983). El cuaterna-rio en la provincia de La Coruña. Modeloselásticos para la formación de cavidades.Publicaciones de la UniversidadComplutense de Madrid, Tesis Doctorales,283 p.

VIDAL ROMANÍ, J.R. (1989).Geomorfología granítica en Galicia (NWEspaña). Cad. Lab. Xeol. Laxe, 13: 80-163.

VIDAL ROMANÍ, J.R. and TWIDALE, C.R.(1998). Formas y paisajes graníticos.Universidade da Coruña, Monografías, 55,411 p.

VIDAL ROMANÍ, J.R. and YEPESTEMIÑO, J. (2004). Historia de la morfo-génesis granítica. Cad. Lab. Xeol. Laxe,29: 331-360.

VIDAL ROMANÍ, J.R., RODRÍGUEZ MAR-TINEZ-CONDE, R. and MACÍASVÁZQUEZ, F. (1984). Evolución de lavertiente granítica de Monte Louro(Galicia, España). Paris, Recherche sur lesCivilisations, Hom. G. Zbyszewski: 101-108.

VIDAL ROMANÍ, J.R., UÑA ÁLVAREZ, E.DE and SANJURJO SÁNCHEZ, J.(2006). Proposal of nomenclatura for gra-nitic forms. IX Reunión Nacional

CAD. LAB. XEOL. LAXE 33 (2008) Description and nomenclature 81

Geomorfología, Santiago de Compostela,Actas: 803-814.

VIDAL ROMANÍ, J.R., GRAJAL BLANCO,M., VILAPLANA, J.M., RODRÍGUEZMARTÍNEZ-CONDE, R., GUITIÁNRIVERA, F., MACÍAS VÁZQUEZ, F.,FERNÁNDEZ SANTÍN, S. andHERNÁNDEZ PACHECO, A. (1979).Procesos actuales: micromodelado en elgranito de Monte Louro, Galicia, España.IV Reunión G.E.T.C., Banyotes-Girona,Actas: 246-266.

VIDAL ROMANÍ, J.R.; TWIDALE, C.R.;FERNÁNDEZ MOSQUERA, D.; UÑA

ÁLVAREZ, E. DE; YEPES TEMIÑO, J.(2004). Rock of Ages. Ann. Bingh.Geomorphology Symp., Kentucky,Abstracts : 31.

WILHELMY, H. (1958). Klimamorphologieder Massengesteine. Brunswick,Westermanns, 238 p.

WILHELMY, H. (1964). Cavernous rock sur-faces (tafoni) in semiarid and arid cli-mates. Pakistan Geogr. Rev., 19: 9-13.

This work has been carried out with theresearch project CGL2004-06516/BTE byScience & Technology Ministry of Spain

CAD. LAB. XEOL. LAXE 33 (2008)82 Uña Álvarez