Devils Tower (Wyoming, USA): a Lava Coulée Emplaced Into a Maar-Diatreme Volcano?

Home , Andesite, Buisson (crater), Columnar jointing, Missouri Buttes, Phonolite, Porphyritic

Závada et al. Devils Tower (Wyoming, USA): A lava coulée emplaced into a maar-diatreme volcano?

P. Závada1,*,†, P. Dědeček1,†, J. Lexa2,†, and G.R. Keller3,† 1Institute of Geophysics of the Czech Academy of Sciences, v.v.i., Boční II/1401, 141 31, Praha 4, Czech Republic 2Geological Institute, Slovak Academy of Sciences, Dúbravská cesta 9, P.O. Box 106, 840 05, Bratislava 45, Slovak Republic 3School of Geology and Geophysics, University of Oklahoma, 100 E. Boyd Street, Sarkeys Energy Center, Suite 710, Norman, Oklahoma 73019, USA

ABSTRACT intrusive body in the form of a magmatic stock primarily on the basis of spatial distribution of (Russell, 1896; Robinson, 1956), a laccolith magmatic fabrics and their kinematics (Cloos We have investigated the mode of emplace- (Pirsson, 1894; Jaggar, 1901), or a volcanic con- and Cloos, 1927; Varet, 1971; Jančušková et al., ment of iconic Devils Tower, which is a phono- duit (Carpenter, 1888; Halvorson, 1980); the lat- 1992; Arbaret et al., 1993; Závada et al., 2009a). lite porphyry monolith in the state of Wyo- ter interpretation is presented in many Earth sci- One of these well-exposed phonolite buttes ming in the western United States. Our fi eld ence textbooks. Although conclusive evidence (Bořeň phonolite body, Czech Republic) was survey of this structure and its geological set- is lacking, it is generally accepted that it repre- recently investigated and interpreted as a rem- ting, its radiometric dating, and the tectono- sents an intrusion of phonolite magma that was nant of a lava dome extrusion into the crater of magmatic evolution of the region suggest a exposed after erosion of the surrounding weakly a maar-diatreme volcano by combined methods new genetic interpretation of the vol cani clastic consolidated Mesozoic and Cenozoic sedi- of magmatic fabrics and cooling fracture pat- rocks in the area and provide a basis for a new mentary layers (e.g., Robinson and Davis, 1995; terns (Závada et al., 2011). hypothetical emplacement scenario for Devils Sigurdsson, 2000). Corry (1988) claimed that The fact that Devils Tower is also associated Tower. This interpretation was inspired by an Devils Tower is a volcanic neck that may have with phreatomagmatic pyroclastic rocks simi- analogy of the tower with a similar phono lite formed from solidifi ed magma that extruded lar in composition to those associated with dia- butte in the Cenozoic vol canic region of the along a failure zone in the extended roof of a tremes elsewhere in the Black Hills (Effi nger, Czech Republic and analogue modeling using Christmas-tree laccolith. Lisenbee and Roggen- 1934; Lisenbee and Roggenthen, 1990; Kirch- plaster of paris combined with fi nite element then (1990) and Rakovan (2006) suggested ner, 1996) and exhibits remarkable similari- thermal numerical modeling. Our results that the emplacement of phonolite magma was ties with the Bořeň body (e.g., size, phonolitic indicate that Devils Tower is a remnant of a somehow associated with diatremes or phreato- composition, inverted fan pattern of columnar coulée or low lava dome that was emplaced magmatic volcanoes; however, they did not jointing) motivated us to undertake a new sur- into a broad phreatomagmatic crater at the address the mode of Devils Tower emplace- vey and analysis in Devils Tower area to con- top of a maar-diatreme volcano. ment. In contrast to all intrusive scenarios, Kiver sider alternative models for its emplacement. and Harris (1999) argued that the tower could We propose a new emplacement scenario that represent a remnant of a surfi cial lava body is supported by a conceptual analogue model INTRODUCTION or welded pyroclastic material emplaced into using plaster of paris as an analogue of magma the phreatomagmatic maar crater. Spry (1962) that shows the internal fl ow pattern from mag- Devils Tower is a dominant landmark of the explained the inverted fan columnar jointing netic fabrics of dispersed magnetic particles northern Great Plains (Wyoming, USA). It rep- pattern on Devils Tower by cooling of an extru- and serves as a template for numerical model resents the world’s fi nest example of columnar sive lava sheet above the conduit based on the of cooling that is matched with the Devils jointing in phonolite and possibly the longest mathematical model of Jaeger (1961). Tower columnar jointing pattern. Owing to the columns developed in a volcanic rock (Fig. 1). Understanding the emplacement mode for National Monument administrative limitations The Cenozoic phonolite porphyry monolith Devils Tower requires additional study using we could not carry out direct structural mea- forming Devils Tower is located in the western modern techniques of structural analysis of surements or collect samples for a systematic Black Hills in northeastern Wyoming (Fig. 2). magmatic fabrics and fracture systems that anisotropy of magnetic susceptibility (AMS) The scientifi c debate on the origin of Devils develop during cooling of magmatic bodies. study as intended. Tower has lasted for more than 100 years. Additional insight can be gained by studies of The majority of previous studies of Devils similar features around the world, such as those CENOZOIC IGNEOUS AND Tower have concluded that it is a remnant of an found in the Cenozoic volcanic provinces of VOLCANIC ACTIVITY IN THE western and central Europe. These features are BLACK HILLS UPLIFT located in the foreland of the Alpine belt (Cajz *Corresponding author. †Emails: Závada: zavada@ig .cas .cz; Dědeček: et al., 1999; Ziegler, 1994), and the emplace- The Devils Tower phonolite monolith (Figs. pd@ig .cas .cz; Lexa: geoljalx@savba .sk; Keller: ment mode of phonolite or trachyte bodies that 1A, 1C–1E) and Missouri Buttes, a group of grkeller@ ou .edu. are similar to Devils Tower has been interpreted fi ve bodies of similar composition located 5 km

Geosphere; April 2015; v. 11; no. 2; p. 354–375; doi:10.1130/GES01166.1; 17 ﬁ gures; 3 tables; 1 supplemental ﬁ le. Received 5 August 2014 ♦ Revision received 18 November 2014 ♦ Accepted 13 January 2015 ♦ Published online 17 February 2015

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/11/2/354/3335355/354.pdf by guest on 03 October 2021 Devils Tower as a lava coulée? A B

WNW ESE C D

UCUC LCLC S B

SE NW NE SW E F DT MB

WEE W

Figure 1. Field photographs of Devils Tower and Missouri Buttes with orientation indicated. (A) Southwestern side of the tower shows vertical and horizontal joints of the base and curved columns plunging to the west. (B) Detail of the southwestern wall shows apparent compositional layering as dark horizontal streaks across the columns (stippled black line), and a suture between the upper and lower colonnades indicated by columns of the upper colonnade that split to two or three narrower columns of the lower colonnade (indicated by arrows). Dashed white lines indicate broad suture between both colonnades. (C) The asymmetrical shape of Devils Tower, with straight columns plunging at ~65° on the southeast and curved columns on the northwest, is seen best from the northeast. (D) In contrast, view from the northwest shows symmetrical shape of subvertical columns on both sides with slight bend to shallower plunge angles on their lower ends. B—base, S—shoulder, LC—lower colonnade, UC—upper colonnade. (E) Missouri Buttes, located ~5 km northwest from Devils Tower, represented by 5 separate phonolite bodies distributed along a periphery of a north-south elongated ellipse of short and long axes of 1 and 2 km, respectively. (F) Two northern buttes of Missouri Buttes in a view from the north.

Geosphere, April 2015 355

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/11/2/354/3335355/354.pdf by guest on 03 October 2021 Závada et al.

103.3°W Black Hills uplift VOLCANICLASTIC DEPOSITS AND 44.9°N STRUCTURAL DEFORMATION N Tertiary igneous rocks OF THE AREA MMBB 50 km Cretaceous DDTT Jurassic(Fm.) Sundance M The distribution of Mesozoic and Tertiary S Triassic and Permian (sFm.)Spearfi h marine and terrestrial sediments surrounding Fig. 3 T MMCC Paleozoic limestones the Devils Tower and Missouri Buttes bodies Precambrian basement (Fig. 3) and the approximate average thick-

River PowderBasin Rapid nesses from borehole data (Robinson et al., City Location of diatremes 1964; DeWitt et al., 1989; Sutherland, 2008) in the Black Hills are presented in Table 1. The youngest exposed MB - Missouri Buttes deposits are represented by volcaniclastic 43.5°N DT - Devils Tower material that is found in four areas (DT-2, S - Sugarloaf MB-C, MB-N, MB-D; Figs. 3 and 4). Photo- Wyoming M - Maitland South T - Tomahawk graphs of some of the samples we collected 104.9°W Dakota 103.3°W MC - Meadow creek are shown in Figure 5: (1) DT-2, found ~230 m west-southwest of Devils Tower is representa- Figure 2. Schematic map of the Black Hills, which represent a tive of rounded clasts of granite, limestone, and lithospheric-scale uplift exposing Precambrian basement in its core limestone conglomerate not found in the sur- and the entire section of the Phanerozoic sedimentary cover. Cenozoic rounding sedimentary units (Fig. 3); (2) MB-C, igneous centers associated by domes in the host rocks are aligned in found in outcrops or blocks in the central area west-northwest–east-southeast direction across the uplift. Stars show between the buttes (Figs. 4B, 5A, and 5B); locations of the six diatremes in the Black Hills (after Lisenbee and (3) MB-N, found between the two northern Roggenthen, 1990). buttes (Fig. 3C) that also contains an erratic granite boulder 1 m in diameter (Fig. 4C); and (4) MB-D, found in two outcrops (~10 × 4 × northwest of Devils Tower (Fig. 1F), represent The maar-diatremes consist of a maar crater, 3 m in size) that reveal fragments of granitic or the westernmost products of Cenozoic igneous surrounding tephra ring, and the underlying dia- metamorphic rocks to 40 cm in diameter (Figs. activity distributed along a northwest-south- treme. The diatreme is fi lled with country-rock 4D, 5C), ~800 m west of the northwestern east–trending belt within the Black Hills uplift and juvenile magmatic clasts and was eventually butte. The outcrops west of Missouri Buttes (Fig. 2). The Black Hills uplift is structurally injected by late-stage intrusions (Lorenz, 2003; display discordant contacts with the underly- the highest segment of a nearly 1000-km-long White and Ross, 2011; Valentine and White, ing Cretaceous Skull Creek Shale. arch that developed due to lithosphere-scale 2012). There are six diatremes identifi ed through- The volcaniclastic rock shown in Figure 4 is folding during the Laramide orogeny (Tikoff out the Black Hills area (Fig. 2) that represent rich in juvenile phonolite clasts and country- and Maxson, 2001) starting ca. 65 Ma (Flores different exposure levels of the original maar- rock clasts from different underlying lithostrati- and Ethridge, 1985; Lisenbee and DeWitt, diatreme volcanoes (Lisenbee and Roggenthen, graphic units and the crystalline basement (e.g., 1993). Precambrian igneous and metamorphic 1990). The shallowest levels are exposed in the shale, limestone, granite, phonolite, and schist) rocks are exposed in the core of the Black Hills Missouri Buttes, Devils Tower, and Sugarloaf that are all encased in a matrix of angular crystal uplift and are overlain by a >1200-m-thick diatremes in the western part of the Black fragments (as large as 3 mm) of quartz, micro- sequence of Paleozoic and Mesozoic sedi- Hills (Fig. 2); deeper levels are exposed in the cline, sanidine, hornblende, aegirine-augite, oligo- mentary strata. Maitland, Meadow Creek, and Tomahawk dia- clase, and fi ne phonolite fragments. The clasts Alkaline igneous rocks in the Black Hills are tremes to the east (Lisenbee and Roggenthen, range in diameter from microscopic to 1 m (Fig. part of the Great Plains alkalic province in South 1990). K-Ar dating of biotite from a pitchstone 4C). Similar volcaniclastic deposits were found Dakota, Wyoming, Montana, and southern embedded in the Tomahawk diatreme indicates during earlier surveys in trench excavations of Alberta (Canada) that originated from parental an age of 55.8 ± 1.4 Ma (Redden et al., 1983); a 150-m-long elliptical knoll, trending west- mantle melts. These melts possibly ascended 40Ar/39Ar dating of sanidines in phonolites southwest from the edge of the main talus on the along the southwest edge of the subducting Kula from Devils Tower and Missouri Buttes pro- west-southwestern side of Devils Tower (Effi n- slab through the Farallon-Kula slab window vides indistinguishable plateau ages of 49.04 ± ger, 1934), and revealed a variety of clasts beneath the North American plate (Duke, 2009; 0.16 and 49.24 ± 0.28 Ma, respectively (Duke similar to that in the Missouri Buttes outcrops, Duke et al., 2014). Differentiated equivalents of et al., 2002). although the matrix was extremely weathered. the Great Plains alkalic province magmas were emplaced in the crust during three episodes (ca. 58 Ma, 55–54 Ma, and 49.6–46 Ma; Duke et al., 2002) in the form of stocks, dikes, and ring Figure 3 (on following page). Geological map and a cross section of the investigated area dikes in the basement, sills and laccoliths in the redrawn after Halvorson (1980) and Sutherland (2008) with Missouri Buttes (MB) and Paleozoic section, and maar-diatremes intruding Devils Tower (DT), and indicating sampling sites and dip marks of the sedimentary strata the entire section (Figs. 2 and 3), including the (masl—meters above sea level). The upper surface of the Fall River datum, redrawn after basement and Paleozoic and Mesozoic strata Robinson et al. (1964), is indicated by solid and dashed red contours that are marked with (Lisenbee and Roggenthen, 1990; Lisenbee and elevations in meters. The cross section reveals an earlier hypothetical interpretation of the DeWitt, 1993). igneous structures (Lisenbee and Roggenthen, 1990; DeWitt et al., 1989).

356 Geosphere, April 2015

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/11/2/354/3335355/354.pdf by guest on 03 October 2021 Devils Tower as a lava coulée?

A MB-NMB-N MB-DMB-D

MB-CMB-C

0 0 4 14001 70 137013

1430 44304

446041460 2 140011430 2 0 4 6 0 4 0 141 0 3939 5252 3 1717 60 5 1414 0 DT-1DT-1 DTDT 141140 6 4 14001 000 40 0 0 0 3 4 DT-2DT-2 141

masl XA B 1500 FallRF. iver m datum ? ? 1000 ? ? ?

0 Vertical exaggeration: 2x Lower Cretaceous Triassic/Permian Quaternary Newcastle Sandstone Spearfish Formation Alluvium Skull Creek Shale Paleozoic Talus Fall River Formation Limestones Stream terrace deposits Lakota Formation (only in section) Precambrian Tertiary Jurassic Granites and gneisses White River Formation (only in section) Analcime phonolite Morrison Formation Redwater Shale and normal fault Foid-bearing Lak Member alkali trachyte Hulett Member inferred normal fault Phreatomagmatic deposits Stockade Beaver Shale Borehole Upper Cretaceous Gypsum Spring Formation _ Mowry Shale Sundance Formation Dip direction mark (only in section) 2 with indicated dip angle

Figure 3.

Geosphere, April 2015 357

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/11/2/354/3335355/354.pdf by guest on 03 October 2021 Závada et al. , e l e s a Local doming of the area is inferred from the e a h c l i s s f t c y

a nonplanar topology of the Fall River Formation o , a o y a e l b r hale, e h a s n l c t s e t i r d r r a s datum (Robinson et al., 1964) that is charac- , n i g d n m o h u e e g f n a k o n s o l n e r o a r m a terized by several domes and depressions. The g p o a a n , e t s g , n o , d e s w t d a e t o s m e r l n o e n ﬁ rst dome, located 2 km southwest of Missouri d f i c s e u i n r l o t i b s n i t n d b d d a i d a r r s i r e n h n

h Buttes, has a diameter of ~1.6 km and a relief o , a y o n a t l t d s a i a e a n a e u l i n s , k n w d s n c c u

f of 90 m; the second dome, located 1 km north i e c m o n , t t d o o n s a e a r e l n s d e o u n t s s b b e t l of Missouri Buttes, has a relief of 60 m (Fig. 3). n s d t l i o o a s u f t l s a e d s t e l s h o s c y i n e , t c r e i d

s The third and fourth domes are characterized by o o b o s s e e a c n t t t t r f n w , h i d n f t o s o n e r a e t r t e o t u d l n m s e t t

a a relief of only ~30 m and are located 2–3 km i n s o l b n e e o s i s , r o e e l n u m l d t h d n e o o a i e g g h s a d p m l to the south of Missouri Buttes and 1 km west n t n i d f h i n a y l e h e a t a o r h o a a t s f n b w t d s h l d i r s h s f

o of Devils Tower, respectively. The last dome s c n g t t y n e t e z k i e l o w t t i t a n i d s a c s y n i r l t w s t n s c l n o u a i o i i a e t o is associated with a phonolite sill exposed in e s r t t r a i t e o d i s s u n n i s e r g s s n s e e y f q o - o e y n r t t n m c o u d o the Barlow Canyon area (4 km northwest of d h i t c h o s d l h o o o m d t i l p n s i s o p e e c g l t s e s e r i l o l a d n l i n , y n t b Missouri Buttes, not shown in Fig. 3) that was m i o a c d , e i e s f i c t a l g n i t a p l e i y m i b h t o y n r l i a d , c n e f r u t r p d e g emplaced under the Hulett Sandstone Member of m s n n e o v - s i o s o e n d d i t i o d h a i n e e l p h o n t i s a n s s s t i l e c t y n t n l o p s a s e l a i e the Sundance Formation (Robinson et al., 1964; e c b i ﬁ w g o r a d h a h n s n t y t a s t h r e f i p t o y o f s y s h o a d , i m a t n t t h r l

, Halvorson, 1980). Drilling in the central part of w a p y i u e a n t e s , r s t o r w i r g u e e t s a n w d e i c e g m g c l l n w s d i d i - u l o - o i n l t o t i l the ﬁ rst dome encountered phonolite porphyry at e t t d a l r l , m l e h o i l l a y i n p s n e l h e e o s n a d s s n k e e e i t e a d s o v v t d r h l l o d i l c i y v v t i s y i 467 m in the Pennsylvanian–Permian Minnelusa l y n w d e , t s e k s o h a a y c s d a l a a y a o v r r i e s p r c l b r e r t i w l n b s h a r a g g b y c r a a g g l r

a Formation (Robinson et al., 1964). These intru- r l e s a l c m e - c o d i d n t g l d d y l e m b a s h o m t n d g e d n r t o n n w i s e e g d d n d o v a n i n

e sions are regarded as comagmatic with Devils e h i a a o i i t v d n a e e e r r , and Sutherland (2008). f r n a i e d n , , , f l e y d d e b n e w e y n y d e m d i d u o e - stones and shaly limestones n a d d d Tower and Missouri Buttes on the basis of com- g t e o a o g e g o a n n b i r h t r r r t o r o e e n o b , z t a y a s g t s l g b g - l a t i o ﬁ b g b k l s s s r t r d r - - o t t k t e d k y positional similarity (Halvorson, 1980). In con- c y a , , r a r e n c e n h e sand and local gypsum beds and fossiliferous oolitic limestone aegirine-augite, nepheline, and analcime h h coal, and bentonite sandstone h bentonite beds phonolite, sedimentary rocks, gneiss, or granite augite, sphene, nepheline, and nosean t t d a i i n t a t c t a l l h l u a r i l a n i i g g g e h h a n i i n i o p i q c s l l d g t b L i l r t ﬁ s s l t trast, a 60 m depression in the Fall River Forma- tion datum surrounds Missouri Buttes. Another depression (~30 m) surrounds Devils Tower and is characterized by a gentle dip of sedimentary strata toward the tower (Fig. 3).

DEVILS TOWER AND MISSOURI BUTTES PHONOLITE BODIES Hulett Sandstone MemberStockade Beaver Shale massive and often cross-bedded sandstone light gray, dark greenish-gray shale Redwater Shale Memberne-grained sandstone glauconitic greenish-gray shale interbedded with light gray ﬁ Lak Memberne-grained sandstone yellowish-gray and nonresistant ﬁ Devils Tower e l a . Devils Tower is a steep-sided phonolite h m S F monolith whose top (elevation of 1558 m above e g h e c

p sea level [asl]) rises almost 250 m above the i e n p n i O underlying sedimentary rocks of the Sundance d W n n Formation (Figs. 1A, 1C, 1D). The base is ellip- d o a i s t n t i s e a a t s e tical with axes measuring 304 m (north-north- i n n e n m e n o s o n n o r t n i o t i n o p i o t o o o o t t i p i s o t e east–south-southwest) and 228 m. The plateau p a t e t i m s n F l a u e t s d e a a o o d a m o l a i e g m d r m r t n h e i r m m o at the top of Devils Tower measures 90 m by n o r e m r a m a i c c L o G S t i r i r D e o o F t l a S y F r L m o p r a k s F F r a d t r h 55 m. Steep columnar jointing characterizes the e d F e e S a a a n o h r o l c e h v l o h e t i s t e p o t n e F o a c a S s r i i o s i u m r l m t a R v o k l upper 180 m of Devils Tower. In contrast, the ﬁ t i w a n a C y w m u r o s u a s t e e i i r i e c e l l a R d s s a a a n l r t t v o n n i m i c w l r a a w u p e o e h l u l r lower section, subsequently referred to as the k u l n n r h h l k o i o i y e e k h l l t p a o a o a a TABLE 1. STRATIGRAPHIC TABLE OF THE LITHOLOGICAL UNITS OVERLYING THE PRECAMBRIAN BASEMENT THE PRECAMBRIAN BASEMENT UNITS OVERLYING THE LITHOLOGICAL OF TABLE 1. STRATIGRAPHIC TABLE D M M L F M W N S G M S P V A S T P F A W base (Fig. 1D), is affected by three joint systems (vertical radial, vertical circumferential, s u

o and horizontal) that divide the cliffs of the base r e f i into blocks ~3 m in size (Dutton and Schwartz, n o s n b

k 1936). The walls of Devils Tower consist of two r a y i c r a c o s s i a r t C

c tiers of columns, like those typically developed v u u i n – s o o o s c e c n i n u d s i e e s r s a a in basalt lava ﬂ ows (DeGraff and Aydin, 1987), s o c c a i m i u i s i s O e a a r o t t a d m m y a r – n T r r r n r r e e e that merge ~70 m below the top plateau (Fig. e g n r r – u e e a a i s u f i i J a n n J P P C C i c y y r i n r a r r 1B). The area above the base, which is also r r r r r i e d v e l o b t a a e e e e e o i o i d b m i t t a p m p r r r w w w d r r d called the shoulder (Fig. 1D), consists of col- i r u p a p a e e o o o e e L U L O U P C T C M T L Q P umns of the lower tier with diameters of 2–3 m.

Lines between rows correspond to unconformities. Compiled after Robinson et al. (1964), Halvorson (1980), DeWitt (1989) These columns taper to widths of only 1.5 m ) 7 4 8 2 0 0 5 5 8 8 5 9 5 0 2 3 7 4 1 1 3 2 5 0 1 6 6 m ( 28 123 60 Upper and Middle Jurassic Sundance Formation 20 15 1 2 2 at the contact with the upper tier, where they Note: Average Average thickness frequently merge into one thick column (> 3 m

358 Geosphere, April 2015

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/11/2/354/3335355/354.pdf by guest on 03 October 2021 Devils Tower as a lava coulée?

A B

DT-2DT-2 MB-CMB-C C D

MB-NMB-N MB-DMB-D

Figure 4. Outcrops of the volcaniclastic rocks in the surroundings of Missouri Buttes and Devils Tower. (A) Rounded block of granite photographed at the DT-2 site. (B) The valley between Missouri Buttes locally hosts blocks of volcaniclastic deposits as long as 2 m. Hammer is 30 cm long. (C) A large rounded boulder of granite (in foreground) was found in phreatomagmatic deposits (in background) at a saddle between the two northern Missouri Buttes. (D) Outcrops of volcaniclastic deposits ~0.8 km west of Missouri Buttes show angular clasts, as much as 30 cm long, of various lithologies encased in earthy brown matrix; a metamorphic rock clast is next to the hammer. The hammer is 30 cm long. Letters in the lower left corner of pictures refer to sampling sites in Figure 3.

wide) of the upper tier (Fig. 1B). Subvertical vertical at the top and plunging ~50° at their 30 vol%), aegirine-augite (6 vol%) zoned to columns of the lower tier ﬂ are outward and lower end, while the columns on the southeast aegirine, sphene (1.2 vol%), and rare amphi- plunge 50°–85° at the shoulder. The thick ver- side are straight, plunging at 65°–75°. Joints of bole. The groundmass consists of anorthoclase, tical columns of the upper tier display bumpy the vertical radial system of the base frequently low albite, K-feldspar, analcime (average whole surfaces with horizontal ledges and abundant merge with the tensional joints of the columns. rock 15 vol%), aegirine, nepheline, and nosean horizontal or irregular cross joints, which are This transition is also locally marked by broad (Halvorson, 1980). Secondary (alteration) prod- accentuated by spheroidal weathering. Blocks joint surfaces curving upward with transverse ucts are represented by calcite, zeolite, hema- that have fallen from the upper part of Devils ribs (Dutton and Schwartz, 1936). tite, clay, and analcime. Vertical or high-angle Tower form the majority of the talus (Dutton plunges (>75°) of magmatic lineations deﬁ ned and Schwartz, 1936). Devils Tower appears Petrography of Devils Tower by alignment of the phenocrysts character- almost symmetrical when viewed from the The analcime phonolite forming Devils ize the bottom part of the tower below 1420 northwest, with the plunge of the columns at the Tower is holocrystalline and coarsely porphy- m asl. In contrast, one sample collected on the shoulder being ~75° (Fig. 1D). From the north- ritic with a gray to olive-gray aphanitic and northeast part of the tower, at an elevation of east (Fig. 1C), the columns of the northwest side trachytic groundmass around phenocrysts of 1450 m, revealed horizontal magmatic lineation of Devils Tower are gently curved, being nearly anorthoclase as much as 16 mm long (average (Halvorson , 1980).

Geosphere, April 2015 359

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/11/2/354/3335355/354.pdf by guest on 03 October 2021 Závada et al. A B C MB-C2

MB-C1 MB-D

Figure 5. Photographs of the sections through samples. Volcaniclastic rocks collected reveal abundant fragments of different lithologies and various sizes. (A, B) Rocks from the central part of Missouri Buttes. Note apparent amoeboidal vesicles in sample MB-C2. (C) Rock from the outcrops 0.8 km west of the buttes. Note the large unweathered phonolite clast in the top part of sample MB-D.

Missouri Buttes trachytic and consists of aegirine and sani- and quartz crystals, and pores. Clast shapes dine laths mixed with isotropic analcime were manually digitized in ArcView GIS (geo- Missouri Buttes consist of ﬁ ve phonolite (Halvorson, 1980). graphic information systems; www.esri.com) and alkali trachyte bodies distributed along the software and statistically processed using the perimeter of a north-south elongated ellipse SAMPLING AND METHODOLOGY PolyLX toolbox in Matlab (Lexa et al., 2005). Φ with axes of ~2 km and 1 km (Fig. 3). The Bulk densities (Db) and porosities ( ) for the highest in elevation is the northwestern butte, A ﬁ eld survey of the area was carried out with volcaniclastic rocks in the Missouri Buttes area reaching 1637 m asl. The three southern buttes a permit from the U.S. National Park Service and the Devils Tower phonolite were measured encompass a triangular elevated plateau and that allowed sampling of sedimentary rocks in and calculated by the triple weighing method; are separated from the two northern buttes by the area around Devils Tower National Monu- weighing the samples dry (A), saturated and a V-shaped valley. Aerial photographs reveal ment, and collection of only one specimen of suspended in water (B), and saturated and that the buttes are affected by steep joint sys- the phonolite in the blocky talus at the base of weighed in air (C), and then calculated using the ρ tems breaking the igneous mass into plates or the tower. This sample was taken from a large following equations: Db = [A/(C – B)] × , and slabs. This joint system strikes at an azimuth column fragment that apparently cleaved off Φ = [(C – A)/(C – B)] × 100, where ρ refers to of 075° in all buttes except for the northeast and collapsed from the upper tier of columns on the density of water. butte (Halvorson, 1980) where this strike is the tower. The axis of the column was recorded The phonolite sample collected from the col- 040°. Margins of the bodies display blocky and marked on the sample. Field surveying and lapsed columnar block at Devils Tower was also jointing forming thin slabs, where radial verti- sampling conducted in the Missouri Buttes evaluated for AMS and bulk magnetic suscepti- cal and peripheral vertical joint systems inter- area focused on the outcrops of volcaniclastic bility variations in the temperature range of 190 sect. From the blocky margin inward, horizon- rocks, their structural position, composition, to 720 °C. The AMS fabric data are presented tal columnar jointing is typically developed, and texture. Three samples of the volcaniclastic using the eccentricity (P), shape (T), and mean

with a maximum column diameter of 1 m. rocks were collected for detailed microscopic bulk susceptibility (Km) of the AMS ellipsoid The jointing also displays portions marked by and backscattered electron (BSE) investigation (Nagata, 1961; Jelínek 1981). The magnetic apparent vertical columns in the central part of and imaging. In addition, two samples from the parameters and methodology are described in the northeast butte, although such columns are Tomahawk diatreme (Fig. 2) were donated by the Supplemental Information (SI) File1. typically four sided and do not show smooth Alvis Lisenbee (South Dakota School of Mines transitions with the surrounding horizontal col- and Technology) for comparison with vol- ANALOGUE MODELING umns. Field observations by Halvorson (1980) cani clastic rocks from Missouri Buttes. These did not reveal any upturned beds that would samples were collected from a layered tuff at a In the second part of this integrated study we suggest warping of the sedimentary layers road cut on the southern edge of the structure employed analogue modeling to constrain pos- around the phonolite buttes. (TD-1) and a lapilli tuff from the central part of sible shapes of magmatic bodies that form by The southwest and west-central buttes the Tomahawk diatreme (TD-2). intrusion of magma into a maar-diatreme with were identifi ed as alkali trachyte, in contrast A Cameca SX-100 microprobe (Department steeply dipping walls fi lled with chaotic vol- to the other three buttes (and Devils Tower) of Electron Microanalysis, Geological Institute caniclastic deposits. Although this scenario is recognized as analcime phonolite. Both of Dionyz Stur, Bratislava, Slovak Republic) types are holocrystalline with an average was used for BSE imaging of selected areas of 1Supplemental Information File. Zipped file phenocryst content of 44%. The phenocrysts polished thin sections. Mosaics of micrographs containing methodology information, including consist of sanidine (20%–32%), anortho- of the volcaniclastic rocks in two thin sections supple mental text, Supplemental Figures SI-1 and SI-2, clase (5%–14%), aegirine-augite (4%–12%), were created for simple modal analysis discrim- Supplemental Tables SI-1 and SI-2, and Movies SI-1 and SI-2. If you are viewing the PDF of this paper or accessory sphene, nepheline, and rare mag- inating the content of lithic fragments (sedimen- reading it offl ine, please visit http:// dx .doi .org /10 .1130 netite enclosed in aegirine-augite. The tary, granitic and/or metamorphic clasts), juve- /GES01166 .S1 or the full-text article on www .gsapubs groundmass is randomly oriented or locally nile magmatic clasts, K-feldspar, plagioclase .org to view the Supplemental Information File.

360 Geosphere, April 2015

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/11/2/354/3335355/354.pdf by guest on 03 October 2021 Devils Tower as a lava coulée?

hypothetical, it is supported by a genetic inter- long surrounded by olive-green groundmass 6F). BSE images revealed that the majority of pretation of the volcaniclastic deposits that is with dark green crystals as large as 2 mm of phonolite fragments are cryptocrystalline and presented herein. The materials and experimen- aegirine augite. The upper face of the column pilotaxitic porous aggregates formed by radial tal setup resemble the instructive experiments and a section cut through the specimen per- to dendritically grown alkali feldspar laths, of the intrusion and erosion of laccoliths in the pendicular to the axis of the column showed some of them with albite rims, between 20 and Black Hills (Jaggar, 1901). The advantage of our lenticular cavities as much as 2 mm long and 500 µm long (Figs. 7A–7C). Some of the mag- analogue modeling approach is that the internal 0.5-mm-thick cavities. A thin section of sample matic fragments show phenocrysts enclosed in fl ow geometry, shape, intensity, and direction of DT-1 revealed that these lenticular and locally a porous mixture of lath-shaped alkali feldspars fl ow in the models can be studied with AMS, triangular, wedge-shaped cavities with irregu- and calcite (Fig. 7D). because we add fi ne magnetite dust to the ana- lar edges occur locally in the vicinity of the Bulk density measurements for samples logue magma that works as a tracer of magnetic porphyroclasts and are preferentially aligned MB-C1 and MB-C2 (Figs. 6B, 6C), revealed fabric in the models. Our analogue modeling is in directions at high angles with respect to the values of 1.874 g/cm3 and 1.539 g/cm3 and not scaled, because the intrusion and/or extru- trachytic fabric in the groundmass (Fig. 6A, corresponding porosities of 28 vol% and 37.5 sion time for Devils Tower and the viscosity of inset). Some of these zeolite-fi lled cavities are vol%, respectively. Results of the fragment the phonolite magma are unknown. designated as miarolitic, because they contain size analysis for the MB samples are presented aegirine-augite needles that grew inward from in Table 3. THERMAL MATHEMATICAL the surrounding matrix. Phenocrysts of feld- MODELING spars are clearly locally dismembered to rotated Tomahawk Diatreme rectangular fragments that are displaced in the The layered tuff (TD-1) from the Tomahawk We created two-dimensional (2-D) fi nite ele- direction of the trachytic fabric. Local altera- diatreme is light brown and contains a mixture ment thermal models of conductive cooling for tion of feldspar to analcime is clearly visible in of microcrystalline juvenile rhyolite fragments various shapes of intrusive and/or extrusive bod- the vicinity of the voids between the fragments with phenocrysts of plagioclase and resorbed ies based on the geometry of intrusion shapes (Fig. 6A, inset). quartz and clasts of Precambrian metamorphic suggested previously and selected analogue rocks (Fig. 6G). There was also a small amount models in this study. Because columnar joints Missouri Buttes of Phanerozoic sedimentary rock and trachyte grow perpendicular to isotherms in cooling igne- Two volcaniclastic rock samples collected at encased in a fi ne-grained matrix (200–500 µm) ous and volcanic bodies (Jaeger, 1961; DeGraff Missouri Buttes (Fig. 3), MB-C1 and MB-C2 of angular quartz, feldspar, and plagioclase and Aydin, 1987), we analyzed the match (Figs. 5A, 5B), revealed decomposed phonolite crystals and biotite. The layers in the sample between the modeled thermal structure and the fragments, as much as 3 cm in diameter, encased are characterized by the constituent fragment observed columnar jointing pattern on Devils in a rusty brown matrix. Sample MB-C2 con- grain size with a maximum diameter of 1 mm. Tower. Thermal models were constructed using tains abundant amoeboidal pores as much as A second sample (TD-2) revealed a similar tex- Comsol and Fracture (Kohl and Hopkirk, 1995) 1 cm in diameter in the matrix. Both samples ture consisting of matrix clasts of 0.5–1 mm software. An initial temperature of 850 °C used contain a great variety of fragments in terms of surrounding fragments of rhyolite as much as in all model runs corresponds to slightly higher size, shape, texture, and composition. In thin 1.5 cm in diameter (Fig. 6H). The microstruc- temperatures than the dry solidus of phono- section, the clasts are subangular to rounded tural characteristics are similar to the Missouri lite magmas (Taylor and MacKenzie, 1975), and consist mostly of phonolite, granite, gneiss, Buttes samples in terms of size and shape of the because we assume that the emplaced magma gabbro, carbonate, dolomite, slate, limonitized fragments, presence of country-rock fragments, was already devolatilized (see discussion of siltstones, sandstone, fi ne-grained pumice (all and abundance of plagioclase, microcline, and Phonolite magma properties herein). In the fi nal average 0.5 mm in size, with maximum size quartz crystals in the matrix. step, a 3-D thermal model was constructed for 1 cm), angular K-feldspar (either anorthoclase The Missouri Buttes samples could be char- the geometry of the analogue model with the or microcline), plagioclase, and quartz crys- acterized as lithic-rich lapilli tuffs due to the best fi t to the corresponding 2-D thermal struc- tals (Figs. 6B, 6C). Amoeboidal vesicles in the abundance of lithic (country rock) fragments, ture and the columnar jointing pattern on Devils matrix of MB-C1 and MB-D samples are partly with the sum of sedimentary and Precambrian Tower. The purpose of this 3-D model was to or entirely fi lled with chalcedony. lithics almost equal to the amount of Tertiary confi rm the results of the 2-D thermal modeling The textures of phonolite fragments range magmatic fragments (Table 2). Both localities and to evaluate the heat budget associated with from aphanitic (Fig. 6B), with rare acicular of volcaniclastic rocks investigated at Missouri emplacement of the magma body. Additional euhedral phenocrysts of feldspar laths and Buttes contain vesicles in the tuffaceous matrix information on the setup of the thermal models aegirine-augite, both randomly oriented around (samples MB-C2 and MB-D); no porosity was and associated thermo physical parameters is the circular pores, to porphyritic types with tra- observed in the fi ne-grained matrix of the Toma- provided in the SI File (see footnote 1). chytic (Fig. 6D) texture of the same minerals hawk diatreme. with elongated pores. Clast shape can some- RESULTS times be described by vesicularity; nonvesicu- Magnetic Fabric and Mineralogy of lar clasts are typically subrounded and smooth Devils Tower Phonolite Petrographic and Microstructural in contrast to vesicular clasts that are angular Description of Collected Samples (Figs. 6B, 6C). Large euhedral crystals con- Variation of the mean magnetic susceptibil-

tained in the phonolite clasts are frequently ity (Km) for phonolite sample DT-1 revealed a Devils Tower crosscut at the edges of the host fragments (Fig. stable decrease with temperature from –200 °C The sample of the phonolite collected 6B). A few phonolite clasts are surrounded by to 0 °C and a distinct drop at ~590 °C, the Curie from the fallen block at Devils Tower (Fig. 3, irregular dark rims (Fig. 6E). One clast contains temperature (Fig. 8A). The susceptibility varia- DT-1) reveals phenocrysts as much as 1 cm a large euhedral feldspar crystal in its core (Fig. tion displayed on this graph (Fig. 8A) suggests

Geosphere, April 2015 361

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/11/2/354/3335355/354.pdf by guest on 03 October 2021 Závada et al.

A B MMB-C1B-C1 Figure 6. Micrographs of the magmatic and volcaniclastic specimens collected at Mis- souri Buttes and Devils Tower. (A) Devils Tower phonolite sample DT-1 shows large pheno- DT-1DT-1 crysts of anorthoclase and A

aegirine augite surrounded by 0 miarolitic cavities. The inset shows a sketch of the micro- 1 mm graph with miarolitic cavities 2mm 1 mm in yellow, feldspar phenocrysts in gray, analci mized feldspar patches as hachured areas, and C p MB-C1MB-C1 D MB-C1MB-C1 aegirine-augites in green. Rose diagram shows the preferred orientation of 40 cavities traced in the thin section cut parallel with K K plane of the mag- 1 3 Gr netic ellipsoid. (B) Three types of phonolite and/or trachyte clasts; a dark aphanitic type at G left, porphyritic with trachytic Dol texture in the center, and vesicular with irregular edges at qtz 2 mm 500 µm lower right. (C) Volcani clastic deposits in the central part of Missouri Buttes reveal a mix- E MB-C1MB-C1 F MB-DMB-D ture of fragment types of angular and subrounded clasts of different lithologies; fragments of phonolite pumice (p), dolomite (Dol), basement granite (Gr), and quartz (qtz). (D) Tra- chytic texture of alkali feldspar laths in a trachyte clast. (E) A clast of phono lite surrounded by apparent dark dust is interpreted as an incipient armored 500 µm 1 mm lapilli. (F) A euhedral anorthoclase crystal with irregular opti- cal extinction is embedded in phonolite pumice. (G) Layered G TD-1TD-1 H TD-2TD-2 structure of sample TD-1 collected at the southwest margin of the Tomahawk diatreme showing igneous and lithic rock fragments. (H) Layered sample from the Tomahawk diatreme reveals large clasts (to 1.5 cm in diameter) of rhyolite with a jagged boundary of the enclosed quartz crystal . 1 mm 1 mm

362 Geosphere, April 2015

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/11/2/354/3335355/354.pdf by guest on 03 October 2021 Devils Tower as a lava coulée?

A MMB-C1B-C1 B pumice MMB-C1B-C1 a-fs ab a-fs a-fs

500 μm 200 μm 1 mm 0.5 mm

MB-C1MB-C1 mag a-fs MB-C1MB-C1 C pumice D cc cc ab an kfs cc

a-fs

200 μm 500 μm 2 mm 2 mm

Figure 7. Backscattered electron images of thin sections. (A) Pilotaxitic porous texture (holocrystalline, in which lath-shaped microlites are arranged in a glass-free felt mesh) of phonolite clast that consists of lath-shaped and zoned alkali feldspars (a-fs) with albite (ab) rims. (B) Similar texture of ﬁ ner crystals as in A is shown for two adjacent clasts. (C) Mixture of albite and K-feldspar (kfs) within another clast surrounded by cryptocrystalline pumice. (D) Another variety of the pilotaxitic alkali feldspar porous groundmass enclosing phenocrysts of anorthoclase (an), globular crystals of calcite (cc), and minute grains of magnetite (mag).

that the magnetic signal is carried by some oxi- the K1 lineation (Fig. 8B), imply an originally fracture pattern that developed during its cool- dized magnetic phase, either titanomagnetite or shallowly dipping magmatic fabric in the upper ing. Here we link all of these considerations in maghemite. Clearly, magnetite can be excluded part of Devils Tower. a discussion evaluating the different hypotheses due to the absence of a Verwey transition (rapid for the origin of Devils Tower and associated

increase of Km at –150 °C; e.g., Tarling and DISCUSSION features previously presented. Our approach of Hrouda, 1993). The AMS stereoplot of the DT-1 combined thermal mathematical and analogue sample (Fig. 8B), where the axis of the sampled The reconstruction of the original shape of modeling was used to constrain the physical column is vertical (star symbol), reveals an volcanic bodies and their emplacement level, conditions controlling the ﬁ nal shapes of inves- angular difference of 20° between the mean speciﬁ cally Devils Tower, requires understand- tigated bodies characterized by distinct patterns

minimum susceptibility direction (K3) and the ing of the geological evolution of the entire of internal magmatic fabrics and cooling joints column axis. In addition, the magnetic linea- volcanic complex, physical properties of the and can be tested in particular for Devils Tower.

tion (K1) and intermediate mean susceptibility magma, erosion estimates, geological setting at We put forward an original hypothesis suggest-

(K2) directions plunge at shallow angles of 18° the time of emplacement, structural deformation ing that Devils Tower could represent a remnant and 8°, respectively. These results, together with of the area, and internal fabrics of the volcanic of a lava extrusion into a maar crater of a pre- alignment of the trachytic fabric subparallel to or magmatic body together with the internal existing maar-diatreme volcano (Fig. 9).

Geosphere, April 2015 363

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/11/2/354/3335355/354.pdf by guest on 03 October 2021 Závada et al.

TABLE 2. MODAL ANALYSIS OF TWO MISSOURI BUTTES AND TWO TOMAHAWK DIATREME SAMPLES Rock fragments MB-C1 MB-D TD-1 TD-2 TABLE 3. GEOMETRICAL AVERAGES OF Tertiary igneous 29.4 29.8 39.2 45.3 EQUAL-AREA DIAMETERS FOR THE FRAGMENTS Sedimentary 19.1 6.4 0.1 0.1 OF THE MISSOURI BUTTES SAMPLES Precambrian basement 11.5 13.0 5.2 5.6 MB-C1 MB-D K-feldspar crystals 3.0 6.6 4.9 4.3 Rock fragments (mm) (mm) Plagioclase crystals 0.8 1.6 8.2 8.6 Tertiary igneous 0.61 0.55 Quartz crystals 2.6 3.3 8.1 7.3 Sedimentary 0.85 0.91 Pores 0.6 Precambrian basement 0.63 0.68 Other minerals 0.3 0.4 K-feldspar crystals 0.42 0.43 Matrix 33.7 38.7 33.9 28.7 Plagioclase crystals 0.59 0.43 Total 100.0 100.0 99.9 100.3 Quartz crystals 0.32 0.33 Note: Rock fragments in percentage. MB—Missouri Buttes; TD—Tomahawk diatreme. Data for TD-1 and TD-2 Pores 0.55 samples are taken from analyses of other samples from the same localities by Kirchner (1996).

Regional Deformation “fossil hill” ~1 km north-northwest of Devils (1901), as well as Efﬁ nger (1934) and Robinson Tower (Fig. 3) could thus be explained by and Davis (1995), interpreted this rock type as The depressions in the Fall River Forma- block tilting in marginal screen rather than an initial portion of emplaced magma that was tion datum (Fig. 3) coincide with sedimentary by folding due to emplacement of a shaft-like rich in entrained country-rock fragments and strata dipping toward both Missouri Buttes intrusion or magmatic stock (Robinson and later formed the marginal part of the intrusive and Devils Tower over a radius of ~1 km Davis, 1995). bodies. Phonolite magma around the entrained and were interpreted earlier to reﬂ ect magma country-rock fragments in the marginal parts of chamber roof subsidence (Halvorson, 1980). Genetic Interpretation of the the intrusions was interpreted to be later decom- An alternative explanation could be marginal Volcaniclastic Deposits posed into an earthy brown rock. Halvorson screen collapse (Fig. 9) around the periphery (1980) described this volcaniclastic material as of phreatomagmatic volcanoes (Lisenbee and Jaggar (1901), who mapped volcaniclastic alloclastic breccia that represents pyroclastic Roggenthen, 1990; Lorenz, 2003; Lorenz and deposits in the Missouri Buttes area as agglom- deposits lining the vents of eroded volcanoes; he Kurszlaukis, 2007). High dip angles of the erates, suggested that they extend throughout also attributed the two volcaniclastic rock out- Redwater Shale Member at what is known as the central area of Missouri Buttes. Jaggar crops 0.8 km west of the western buttes (MB-D;

A B P = 1.019 600 N T = –0.017 Kb [E-6] 500

400

270 90 300

200

100 K1 180 K2 0 K3 –200 0 200 400 600 800 T[°C ]

Figure 8. (A) Thermomagnetic curve of the sample DT-1. Kb is the bulk susceptibility of the sample measured (Nagata, 1961). (B) Corresponding lower hemisphere equal-area stereoplot of the anisotropy of magnetic susceptibility (AMS) measured from six cubic specimens. Axis of the sampled phonolite column is marked by the star symbol. P and T correspond to the eccentricity and shape, respectively, of the AMS fabric (deﬁ nitions in the SI File [see footnote 1]).

364 Geosphere, April 2015

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/11/2/354/3335355/354.pdf by guest on 03 October 2021 Devils Tower as a lava coulée?

2011); (12) matrix-supported nature of the Diatreme breccia (Fig. 5), typical for phreatomagmatic breccia pipes with reiterated breccia formation pulses by hydrovolcanic explosions (Tȃmaș 0 km and Milesi, 2002); and, (13) similarity of the rocks we investigated to the Tomahawk dia- MB treme samples (Figs. 6G, 6H) that were inter- DT preted as phreatomagmatic deposits (Lisenbee S and Roggenthen, 1990; Kirchner, 1996). 1 2 In agreement with Lisenbee and Roggenthen (1990) and Kirchner (1996) for the Tomahawk 1 km MC diatreme, we suggest that the infl uence of gas pressure release from the magma had only a T minor effect on the fragmentation of the magma. However, juvenile phonolite and/or trachyte fragments with internal vesicles and irregular Tertiary intrusive rocks (phonolites) boundaries indicative of gas pressure release– Debris jet deposits driven fragmentation can be found locally (Fig. 2 km Diatreme deposits with xenoliths Cretaceous 6B), but vesicular clasts are typically found in Jurassic diatreme deposits (Ross and White, 2012). Permo-Triassic Spearfish Fm. 3 Mississippian Pahasapa Limestone Cambrian Deadwood Formation Erosion in the Black Hills and Precambrian basement Emplacement Level for Phonolite Magma 1 - ring collapse 2 - marginal screen Redrawn after The area of the Black Hills was part of a fore- Lisenbee and Roggenthen (1990) land basin that contained a 2000–2500-m-thick 3 - plutons sequence of Phanerozoic sediments (Lisen- bee and DeWitt, 1993) prior to uplift during Figure 9. Schematic sketch of a diatreme (redrawn after Lisenbee and Roggenthen, 1990). Laramide orogenesis. The combined thickness Erosion levels for respective diatremes as suggested by Lisenbee and Roggenthen (1990) are of eroded layers from the Mowry Shale (young- indicated by dashed lines (for locations and abbreviations, see Fig. 2 legend). est Cretaceous member exposed at Missouri Buttes) to the Fox Hills Formation in the Black Hills monocline ~10 km west of the map area Fig. 3) to horizontal pyroclastic sills (tuffi site; phonolite fragments typical for transport in is ~1850 m (Halvorson, 1980; DeWitt et al., Reynolds, 1954). fl uidized particulate jets in the craters, where 1989). Considering the: (1) widespread and As suggested by earlier studies of similar rounding is also a result of clast recycling (e.g., rapid erosion of the weak muddy upper Cre- volcaniclastic rocks, our fi eld survey and micro- Lorenz and Kurszlaukis, 2007) and milling taceous foreland sediments in the Black Hills structural observations of the collected samples (e.g., Campbell et al., 2013) during fragmenta- before emplacement of Devils Tower (Smith show that the volcaniclastic rocks in the vicin- tion (Figs. 6B, 6C, 6E); (6) blocky and poorly et al., 2008; Fan et al., 2011); (2) a suffi cient ity of Missouri Buttes and Devils Tower repre- vesiculated juvenile magma fragments (Figs. time span (16 m.y.) for erosion between the sent phreatomagmatic deposits (Lorenz, 1973, 6B, 6F) (Morrissey et al., 2000); (7) fi ne pilo- onset of the Black Hills uplift to the emplace- 1986, 2003; Fischer and Schmincke, 1984; taxitic or trachytic texture of these fragments ment of the phonolite bodies; and (3) discordant Lisenbee and Roggenthen, 1990; Martin and (Figs. 7A–7C) indicating rapid undercooling at contact of phreatomagmatic deposits at Mis- Németh, 2005; Konečný and Lexa, 2003; Auer the time of fragmentation due to a contact with souri Buttes with underlying Cretaceous sedi- et al., 2007; White and Ross, 2011; Závada external water; (8) magmatic clasts containing ments (Fig. 3), the phreatomagmatic deposits et al., 2011). This interpretation is supported large phenocrysts that are cut at clast margins, 1 km west of Missouri Buttes can be interpreted by the: (1) variety of fragmental accidental refl ecting high energy of the fragmentation as remnants of surface deposits, possibly a tuff lithic clast types (country-rock clasts picked mechanism (Fig. 6B); (9) low densities of the ring surrounding the original maar. The proxim- up by the pyroclastic jets) with fragments from volcaniclastic rocks in contrast to surround- ity of the present exposure level to the Cenozoic lower stratigraphic levels including crystalline ing sediments (Schulz et al., 2005; Everson paleosurface in the western part of the Black rock basement; (2) relatively large amount of and Roggenthen, 1988); (10) clasts with nar- Hills is also supported by surface volcanic facies fragmented accidental lithic clasts with almost row dark rims that can be interpreted as incipi- of welded ash tuff (Fashbaugh, 1979; Halvor- equal proportion as the juvenile magmatic clasts ent armored lapilli (Fischer and Schmincke, son, 1980) at Sundance Mountain, 33 km south- (White and Ross, 2011); (3) great clast size 1984), where fi ne-grained dust agglutinated east of Devils Tower. variety ranging from microscopic fragments on the host fragment (Fig. 6E); (11) hollow or to boulders 1 m in diameter (Figs. 4C and 6E); chalcedony-fi lled pores in samples from the Analogue Modeling and AMS Analysis (4) large proportion of small particles (Fig. 6) central part of Missouri Buttes (MB-C2) that suggestive of high fragmentation energies may refl ect entrapped volcanic gasses or steam The emplacement of the Devils Tower lava (Valentine et al., 2014); (5) rounded and partly in the deposited muddy lapilli tuffs (Fig. 5B) body is, in our hypothetical scenario, consid- faceted shape of accidental lithic and cognate (e.g., Scarpati et al., 1993; White and Ross, ered as the fi nal stage of phreatomagmatic

Geosphere, April 2015 365

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/11/2/354/3335355/354.pdf by guest on 03 October 2021 Závada et al.

maar-diatreme volcano evolution, when the A infl ux of water triggering the phreatomagmatic explosions in the root zone of the volcano clay and sand layers (maar sediments) pressure ceases, either due to change in hydrological LOAD B control regime (drop of water table) or, less likely, the increase of lithostatic pressure in the deeper root zone of the diatreme (Konečný and Lexa, 2003; Lorenz, 2003; Martin and Németh, 2004, 2005; chunks Auer et al., 2007; Valentine and White, 2012). of solid E To evaluate the similarity of the experimental plaster models created by intrusion of model magma + (plaster of paris) into model diatremes with a sand simple hydraulic squeezer apparatus (Fig. 10) C and the original (Devils Tower), the shape of overpressured D Devils Tower and its relative dimensions in the plaster hypothetical maar-diatreme crater were pro- jected to fi t within the vertical sections of the Figure 10. Experimental apparatus used for analogue modeling of sliced analogue models (Fig. 11). magma injection into maar-diatreme structures. The apparatus Although the crater dimensions of the hypo- consists of a hydraulic squeezer (A), a steel frame transferring thetical diatreme under Devils Tower are not the load of the squeezer (B) on a rectangular steel plate with a known, the typical maximum depths of 1–2 km central circular aperture, a container for analogue magma (C), would correspond to craters 1–1.5 km wide a steel conduit attached to the steel plate (D) above the aperture, (Lorenz, 2003; Valentine and White, 2012). The and a cone-shaped crater fi lled with analogue diatreme debris (E). vertical scaling of the original body that formed During the experiment, the analogue magma (plaster of paris) Devils Tower in our modeling is constrained by evacuates from the container by the force of the squeezer through the level of the suture between both colonnades the conduit and intrudes the analogue debris in the cone-shaped of columns that formed during cooling of the crater, i.e., the diatreme. A shallower maar crater atop the model body. This suture should correspond to the half- diatreme was partly fi lled with alternating sand and clay layers to height of the original body, if we assume that simulate the sediments of the maar in some of the experiments. For the cooling gradients at the top and bottom mar- additional explanation of the experimental setup, see the SI File (see gins of the body were similar. In the second step, footnote 1).

mixing ratio (MR): 2.5 D-11 D-9 D-4 D-2 D-5 D-14 D-7 D-12

D-8 D-6

10 cm 10 cm 10 cm10 cm 10 cm

mixing ratio (MR): 2.15–2.4 D-19 D-18

D-16 MR=2.15

10 cm 10 cm 10 cm 10 cm D-17 D-24 D-15 D-2 2 D-23 D-21 MR=2.2 MR=2.4 MR=2.24 MR=2.4 MR=2.4 MR=2.3 MR=2.3 MR=2.2

Figure 11. Shapes and internal fabrics redrawn from vertical sections of the experimental bodies. Note that outlines of similar experiments are displayed as superposed on each other and are marked by contours of different colors. Mixing ratio corresponds to the weight proportion of hemihydrate powder and water used for preparation of the plaster of paris in the experiments.

366 Geosphere, April 2015

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/11/2/354/3335355/354.pdf by guest on 03 October 2021 Devils Tower as a lava coulée?

for the selected analogue models that fulfi ll the magma overpressures for their buildup (D-6, exceeding the present aerial extent of the tower” above defi ned dimensional preconditions, fi nite D-14); (3) displayed stems that were too wide (Russell, 1896; Robinson, 1956, p. 13); (2) a element thermal models of cooling were cre- (D-22); or (4) revealed unrealistically infl ated volcanic conduit that transferred the phonolite ated to further test the similarity of the thermal edifi ces (D-11) that emplaced rapidly (3 s) magma to higher levels of an eroded volcano structure of the models with columnar jointing after reaching relatively high magma overpres- (Carpenter, 1888; Halvorson, 1980); or (3) a pattern on the original (Devils Tower). sures. The original shape for Devils Tower can laccolith, emplaced approximately between Of the total 27 analogue experiments that be roughly estimated considering that the joints the Jurassic and lower Cretaceous sedimen- were created with the plaster of paris as the confi ning the columns grow perpendicular to tary layers (Pirsson, 1894; Jaggar, 1901). The analogue magma, 10 are shown in Figure 11; the margins of original volcanic bodies during last model (Fig. 14D), based on the results of 7 models shown are overlain with the outlines conductive cooling (Jaeger, 1961; DeGraff and our study, is a large lava coulée emplaced into of 9 additional models that were very similar in Aydin, 1987). The latter intuitive reconstruc- the maar crater of a phreatomagmatic volcano. geometry. The shapes of the experimental bod- tion suggests a laterally (rather than vertically) The shape of the last model mimics the ana- ies varied primarily due to different thickness extensive intrusion or extrusion with relatively logue model extrusion D-19 (Fig. 12, Movie of the plaster defi ned by mixing ratios (MR) fl at surface at the top and a base that dips toward SI-1 [see footnote 1]) with the fl at base of this of the hemihydrate powder and water, initial the vertical axis of the body (Pirsson, 1894; model positioned at the base of the tower within overpressure, and setup of the experimental Jaggar, 1901; Spry, 1962). a hypothetical 1-km-wide crater at the level of maar-diatreme (Fig. 11; Fig. SI-1 and Table One of the models (D19; Figs. 11 and 12) that the Sundance Formation that tapers toward the SI-1 in the SI File [see footnote1]). The details revealed an excellent match of its thermal struc- underlying rock formations. of the experimental runs and movies capturing ture from preliminary fi nite element thermal Inspection of the intrusive models (Fig. 14) some of the experiments are presented in the models of cooling (described in the following) with superposed lines drawn perpendicular to SI File (see footnote 1). In summary, experi- with the columnar jointing pattern on Devils the isotherms and profi le of Devils Tower scaled ments created with relatively dense plaster Tower was further investigated in detail for to match the extent of the intrusions revealed (MR ~ 2.5) required relatively high overpres- internal AMS fabric. This model could be char- that: (1) for a magmatic stock (Fig. 14A), colum- sures of the squeezer (to 130 bar) to form tear- acterized as a lava coulée or low lava dome that nar jointing should converge radially inward, drop-shaped cryptodomes (D-6, D-14), small fi lled the bowl-shaped crater of the maar (Figs. creating a rosette pattern of columnar jointing in extrusive domes with narrow stems (Fig. 11; 12A, 12B) and was fed by a narrow conduit that vertical cross section (Spry, 1962); (2) the cool- Table SI-1 [see footnote 1]; D-5, D-6, D-7, D-8, penetrated along the walls of the experimental ing of the volcanic conduit (Fig. 14B) should D-14), or asymmetric extrusions that escaped diatreme. The conduit revealed bifurcating pro- develop two tiers of horizontal columns in verti- from under a tilted lid of maar-sediments (D-9, trusions and local corner fl ow (Figs. 12C, 12D), cal cross section or a chevron pattern of colum- D-12). A few initial experiments that were cre- where the intruding plaster slightly uplifted the nar joints that are slightly bent upward to form a ated quickly (5 s) at moderate initial overpres- loosely packed fi lling of the diatreme. The fl ow cusp at the vertical contact suture (Spry, 1962), sures resulted in asymmetric extrusions with in the conduit was then channeled into a thin (3) cooling of a laccolith above the conduit thick stems (D-1, D-2, D-3, D-4; Table SI-1 tube-like neck 10 cm below the extrusion. The (Fig. 14C) could produce two fans of columnar [see footnote 1]; only D-2 and D-4 are shown disrupted color banding pattern of the model in joints; one growing and converging from the in Fig. 11). the vertical section (Fig. 12D) shows an onion- base upward, where the conduit fed the laccolith For the second series of experiments with peel–like internal fabric above the conduit and with magma (inverted fan), and a second one more diluted plaster (MR ~ 2.15–2.4), the concave fl ow planes facing the margins of the growing in the opposite direction from the con- experiments revealed similarly shaped asym- body. The AMS fabric of this body revealed a cave roof. Comparison of the columnar joint- metric coulées (D-16, D-17, D-24), low lava pattern similar to our earlier experiments on ing patterns resolved from the thermal models domes with thick stems (D-15, D-22), or rela- coulées and low extrusive lava domes (Závada for the three intrusive geometries and that of tively symmetrical low lava domes or coulées et al., 2009b). The magnetic foliations regularly Devils Tower (Figs. 14A–14C) reveals a rough with narrow feeding conduits (D-19, D-21, dip at high angles to the inspected vertical sec- match only for the laccolith. However, for lac- D-23). All these were created at relatively low tion, and their trends conform to the stretched coliths that have concave roofs, we expect that loading pressures of the squeezer (<80 bars). In and disrupted color banding in the model, defi n- the upper colonnade columns would converge one experiment, the feeding conduit in the dia- ing fl attened concave fl ow planes verging to the slightly downward, but the opposite is displayed treme apparently bifurcated after creation of the margins of the experimental coulée (Fig. 13). at Devils Tower. fi rst dome to produce another dome at the oppo- The thermal model created for our experi- site side of the maar (D-18; Fig.11). A few of Columnar Joint Patterns Constrained by mental extrusion D-19 (Fig. 14D) reveals a the models had to be discarded for later analysis Thermal Numerical Modeling of Cooling nearly perfect match between the resolved due to leaks in the injection system during the columnar jointing pattern (inverted fan; Spry, experiment (D-10, D-13) or rapid burst of the Thermal numerical models of cooling were 1962) and columnar jointing on Devils Tower magma, when the inlet was stuck or the intru- created for four geometries representing differ- just above the fl aring conduit, where the fl ow sion conduit in the diatreme suddenly expanded ent emplacement scenarios of Devils Tower to extruded laterally. Cooling of the coulée from or bifurcated (D-11, D-20). test the match between the thermal structure of its top fl at surface would create vertical columns For the thermal mathematical modeling, the the models and columnar jointing pattern on the growing inward, where they would meet with above-defi ned dimensional preconditions for the tower. Three of the geometries, suggested previ- the curved columns of the inverted fan colon- models excluded a majority of the experiments ously, represent intrusions of phonolite magma nade. In contrast to the laccolith, the fact that the (Fig. 11), because they were: (1) performed too into Jurassic–lower Cretaceous strata (Fig. 14), columns on Devils Tower converge toward the quickly, in 5–10 s (D-2, D-4); (2) displayed specifi cally: (1) a magmatic stock of unspecifi ed top of the tower could refl ect a rather shallowly shapes of cryptodomes that required high shape with horizontal dimensions “not greatly convex or sagged top surface in the central part

Geosphere, April 2015 367

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/11/2/354/3335355/354.pdf by guest on 03 October 2021 Závada et al.

A F B B

B F

C F D

F B

Figure 12. Views of lava coulée experiment D-19. (A, B) Oblique top views. F and B refer to the front and back sides of the model (see text). Scale bar in A and B is 10 cm long. The white dashed line in A, B, and C indicates the trend of the vertical section presented in D. (C) Bottom view. (D) Vertical section photograph. Note the spirally radiating shear zones on the surface of the model (A and B), the irregularly shaped conduit, where the plaster penetrated the model diatreme (C), and ﬂ at bottom part and shallowly inclined margins of the model, where the plaster spilled into the maar (C and D).

of such an extrusion. Spry (1962) also explained Emplacement Mode of Devils Tower further imply that their emplacement level and the inverted columnar jointing pattern on Devils evolution of their magmas or emplacement time Tower by cooling of an extrusive lava sheet The volcaniclastic deposits found at Missouri were similar. It is thus logical to assume that above the conduit; however, in his view, the Buttes are traditionally regarded as analogues to both Missouri Buttes and Devils Tower could be tower would represent a remnant of only the deposits found in the vicinity of Devils Tower interpreted as magmas emplaced at shallow lev- lower half of such a sheet, interpreting the entire (Efﬁ nger, 1934; Robinson, 1956; Halvorson, els of two individual maar-diatreme volcanoes. columnar section of the tower as the lower col- 1980). The exotic clasts (e.g., granite, limestone Furthermore, summarizing the ﬁ eld evidence onnade that formed by cooling from the base of conglomerates, etc.) found at the DT-2 local- and genetic interpretation of the volcaniclastic the sheet. ity (Fig. 3) can thus be explained as material deposits in the area together with the analogue The 3-D thermal model that was constructed ejected from lower stratigraphic levels to their modeling results based on the hypothetical for the geometry of model D-19 also revealed present position by the phreatomagmatic explo- reconstruction here (Figs. 11 and 14), we sug- that the total cooling time in the center of the sions. The discordant contact of the phreato- gest that Devils Tower could represent a rem- hot core at the suture between the cooling fronts magmatic deposits west of Missouri Buttes and nant of an extrusion, i.e., a lava coulée or low above the feeding conduit (Fig. SI-1 [see foot- almost identical ages for both Devils Tower and lava dome into a maar of a maar-diatreme vol- note 1]) corresponds to ~4000 yr. Missouri Buttes, ca. 49 Ma (Duke at al., 2002), cano. Alignment of diatremes along faults is a

368 Geosphere, April 2015

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/11/2/354/3335355/354.pdf by guest on 03 October 2021 Devils Tower as a lava coulée?

Figure 13. Diagrams showing the magnetic

S S

S S S

fabric of model D-19 contoured by the A S S

S S

S S S

S S S S S

S S S

S S S S S S

S S S

S S

S S S S

S S

inverse distance weighting method using S S S

S S S S S

S S S S

S S

0–10 S S S S S S S S

S S S

S S S S S S S S

S S S

S S S S

S S S

S S S S

S S

S S S S

S S

the ArcMap 10.1 software (www.esri.com). S S S S S

S S

S S S S S S S S S S

S S S

10–20 S S S S S S

S S S S S S

S S S

S S

S S S S

S S

S S S S

S S S

S S

S S S S

S S S S S

S S S

S S S Vertical section of the model was drilled in a S S S S

S S S

S S S S S

S S

20–30 S S S S S

S S S S

S S S S S S S S S S S

S S

S S S S S S S S S

S S S

S S

hexagonal grid of 1 cm with cores 0.8–1 cm S S S S

S S S S S S S S S S

30–40 S S S

S S

S S S S S S S S

S S

S S S S S S

S S S

S S S S

long and 1 cm in diameter providing total S S

S S S 40–50 S

S S S

S S S S

335 anisotropy of magnetic susceptibility S S S

50–60 S S S

S S

(AMS) samples. A northwest-southeast S S

60–70 S

S S

elevation profile across Devils Tower is S

70–80 Plunge angle of AMS lineations S S 10 cm S

superposed over the diagrams to illustrate 80–90 from vertical section S S

the match of inspected magnetic fabric with

o o

magmatic fabrics on Devils Tower (Fig. 15). o o

o o

o o o

o o

o o o o

o o o

o o

o o o o o o

o o o o o o o o

B o o

o o o (A) Plunge angle of magnetic lineations o o

o o

o o o

o o

o o o o o

o o o o o o

o o

o o o o

o o o o o

o o o

o o

o o o o

o o o

o o

o o o

50–60 o

o o

from the vertical section shows lineations o o o

o o

o o o o o o o

o o

o o o

o o

o o o

o o o o o

o o o o

o o

o o o o o o

o o o

o o o o o o o o

o o

o o o o o

o o o o

o o

60–70 o o

o o o o perpendicular to the vertical sections in o

o o o o o o o

o o

o o o o

o o o

o o o o

o o o o o o

o o o o o o o

o o o o

70–80 o

the upper part of the extrusive lobes, o o o o

o o o o

o o o o o

o o o

o o o o

o o o

o o

o o o

o o o o o o

80–90 o o o o

in contrast to the base of the model and o

o o o

o o

o o o

the domain above the conduit, where these o

o o o

o o

lineations are parallel to the vertical section. 3 1 3 o o

o o

(B) Arrangement of magnetic foliations o

Dip angle of AMS foliations o

o matches with the pattern of disrupted color from vertical section o o 2 o banding on superposed vertical section of the model (Fig. 12D). Dip angles of all the magnetic foliations from the vertical section C 1.34457 regularly show high values of 80°–90°. Thick red dashed line in A and B indicates a level where vertical fabrics in the walls 1.30044 of the tower pass to exclusively horizontal fabrics in the upper part of the model. 1.27863 1.25683 (C) The contour diagrams of the magnetic 1.23502 fabric intensity. (D) The shape of magnetic ellipsoid. The domain above the conduit P-parameter of AMS fabric (domain 1) is characterized by relatively 1.06416 (fabric intensity) strong oblate fabric (P = 1.20–1.33) with lineations that plunge at low angles from the vertical section. The base of the coulée D (domain 2) is defi ned by an imbricated fabric 0.987988 0.808585 pattern consistent with lateral fl ow of the 0.632632 analogue magma from the feeding conduit 0.456679 0.280726 and shows low intensity (eccentricity of the AMS fabric, P = 1.2–1.28) and moderately oblate shapes (shape of the AMS fabric, T = 0.2–0.82). Fabrics in the coulée above the –0.839821 T-parameter of AMS fabric base (domain 3) can be characterized by (shape of fabric) continuously changing shape of the fabrics from strongly prolate in shallow levels to strongly oblate in the deeper parts of the fl ow with magnetic lineations plunging at high angles from the vertical section. The transition between the base and interior of the coulée (the lateral fl ow) is marked by a narrow band of relatively intense fabrics (P = 1.3–1.32) that are highly oblate close to the conduit and moderately oblate on the frontal side of the coulée (Fig. 12D). A similar, although less expressive, pattern of fabric orientation and shape is revealed on the back side of the extrusion, where the fl ow was blocked by the walls of the crater.

common feature of magma-water interaction the maar-diatreme structure and surrounding at Missouri Buttes. This level is interpreted as processes established in hard-rock environments host rocks. However, this seems improbable the paleosurface at the time of Devils Tower represented by limestones or granites (Lorenz, due to the rather shallow level for the base of emplacement. Furthermore, it is difﬁ cult to con- 2003; Auer et al., 2007; Závada et al., 2011). Devils Tower in the succession of surround- sider that a laccolith would be emplaced at the Alternatively, Devils Tower could be also ing rocks and the reconstructed maar-diatreme currently exposed level, because emplacement explained as a remnant of a laccolith (Pollard and volcano; the recent top surface of Devils Tower of sills or laccoliths is likely controlled by rigid- Johnson, 1973; Corry, 1988), emplaced within matches the level of phreatomagmatic outcrops ity contrasts of the host-rock strata rather than

Geosphere, April 2015 369

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/11/2/354/3335355/354.pdf by guest on 03 October 2021 Závada et al.

A °C B 700

MAGMATIC 600 VOLCANIC STOCK CONDUIT 500

400

300

200

100

C °C D 700

600

500

400 THIS STUDY: MODEL D-19 300

200

LACCOLITH 100

Figure 14. Two-dimensional thermal models of cooling for intrusive bodies based on analogue model D-19 (this study). (A) A magmatic stock. (B) A volcanic conduit. (C) A laccolith. (D) A lava coulée. The geometry of the laccolith was calculated using the equations of Pollard and Johnson (1973). Thin black lines drawn perpendicular to the isotherms (thin white lines) correspond to the columnar jointing that should develop during cooling of the bodies. Thick dotted white lines indicate the sutures between the colonnades. Black dashed lines evince the internal fabric. White thick contours correspond to the northeast-southwest symmetrical elevation profi le across Devils Tower for A, B, and C; northwest-southeast profi le is superposed only for the lava coulée model in D (for an explanation, see text discussion of the asymmetry and erosion of Devils Tower; Fig. 16). Temperature distribution indicated corresponds to 400 yr of cooling after emplacement of a phonolite body with initial temperature of 850 °C. The layers in the model correspond to the major lithologies in the stratigraphic column, excluding the basement. From bottom to top these layers correspond to early Paleozoic limestones, Permian–Triassic Spearfi sh Formation (claystone and/or siltstone), Jurassic Sundance Formation (sandstone), lower Cretaceous sediments (sandstones and/or claystones), and upper Cretaceous shales (e.g., Mowry Shale and younger and eroded Belle Fourche Shale). Scale bars are 500 m. Internal fabrics in the volcanic conduit (A) show Poiseuille-type of magma fl ow generating magmatic foliations that are parallel to the margins, although in the middle part of the pipe, the fl ow planes are horizontal (e.g., Závada et al., 2009b). For both plutons and laccoliths, the latter pattern is modifi ed by the lateral divergence of the fl ow (e.g., Paterson et al., 1998; Buisson and Merle, 2002; Kratinová et al., 2006; Závada et al., 2009b; our personal observations), where the conduit extends into the thickened bulb of the pluton or the laccolith.

by the level of neutral buoyancy (Kavanagh (1988) hypothesis that Devils Tower could rep- are frequently fed by dike-like bodies growing et al., 2006). This assertion is supported by the resent a neck formed from magma that invaded as ﬂ uid-ﬁ lled cracks (Corry, 1988; Price and fact that the majority of the laccoliths and sills in the extended roof of a laccolith can be dis- Cosgrove, 1990) rather than cylindrical pipes. the Black Hills formed by phonolites are found carded, because the surrounding strata are not Cooling of a laccolith fed by such a dike would in the Paleozoic–Mesozoic section (Halvorson, dipping away from this structure, as would be thus result in a so-called “tower” that is elon- 1980; Lisenbee and DeWitt, 1993). Corry’s expected for this scenario. In addition, laccoliths gated horizontally (a wall).

370 Geosphere, April 2015

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/11/2/354/3335355/354.pdf by guest on 03 October 2021 Devils Tower as a lava coulée?

Internal Fabrics of Devils Tower Devils Tower further reveals that the top part fi lling was relatively high. We presume that the of the tower can be associated with horizontal subvertical joint sets of the base formed per- Sample DT-1 (Fig. 6A), which could only oblate and relatively strongly anisotropic fab- pendicular to the original base of the extrusion. be oriented with respect to the axis of col- ric, which implies uniaxial deformation and/or Smooth transitions at the shoulder of the tower lapsed column, implies originally subhorizon- fl ow of magma with vertical maximum com- between the vertical joints of the base and the tal fabrics (Fig. 8) for the upper part of Devils pressive stress. This is compatible with the joints composing the wide columns of the lower Tower, in agreement with the data of Halvorson microstructure of the DT-1 sample that reveals colonnade suggest that some of the incipient (1980); subhorizontal above the elevation of phenocrysts dismembered and stretched along tensional joints of the lower colonnade were ~1460 m asl and subvertical in the lower part the trachytic fabric (Fig. 16B), and abundant established on the orthogonal joints of the base (Fig. 15). Horizontal fabrics at similar heights miarolitic cavities that are aligned preferen- and that both blocky and columnar joint systems on the southwest side of the tower can be also tially at high angles to the general trend of formed at similar cooling conditions. The thin- refl ected by the possible compositional layering trachytic fabric in the groundmass (Fig. 6A). ning-upward columns of the lower colonnade that is visible approximately at the level where These cavities form due to strain heterogene- on Devils Tower suggest that the rate of cool- the thick columns of the upper colonnade split ities in the vicinity of phenocrysts in the phono- ing associated with formation of these columns to two or three columns in the lower colonnade lite, and refl ect the shear thickening rheology of increased during solidifi cation of the entire (Fig. 1B). the material that extrudes at low confi ning pres- magmatic body. This can be possibly explained Closer inspection of the internal fabric map sures (Smith, 2000; Smith et al., 2001; Závada by a faster solidifi cation due to degassing driven for analogue model D-19 with superposed et al., 2011). Therefore, although formation of crystallization (Sparks et al., 2000), while some profi le of Devils Tower (Fig. 13A) reveals that these cavities cannot be excluded for any of the portion of the gasses that escaped from the crys- onion-peel like arrangement of fl ow planes scenarios considered (Fig. 14), their abundance tallizing magma circulated in the columnar frac- above the fl aring feeding conduit would easily should in general increase with decreasing ture system of the lower colonnade between the explain vertical fabrics at the bottom part of the emplacement depth. base of the body and the solidus envelope (Fig. tower and exclusively horizontal fabrics from 16B), providing faster convective cooling of the about its half-height upward, although similar Cooling Dynamics of the Original lower colonnade (DeGraff and Aydin, 1993). fabric patterns also characterize the intrusive Magmatic Body For the upper colonnade, fl uid content of lava scenarios, i.e., stock, volcanic conduit, and lac- is low (due to low pressure) and any exsolved colith (Fig. 14). The bottom boundary of the The mutually orthogonal sets of columnar gasses would escape into the atmosphere, which domain represented by horizontal fabrics in joints in the base of the tower (vertical radial and is compatible with stable width of the upper the thermal model for D-19 (Fig. 14D) roughly vertical peripheral joint sets) can be explained colonnade columns. Even more likely, the fl u- corresponds to the suture between both colum- to form by cooling at the base of the original ids providing the convective cooling could have nar jointing colonnades. The comparison of the extrusion, when the thermal gradient between been released from the wet base of the original fabric pattern for model D-19 (Fig. 12B) with the hot extrusion and relatively cool diatreme maar buried by the extrusion. Faster cooling of the lower colonnade would also explain why the suture of both colonnades is at about two thirds of the Devils Tower height, if we assume that the present summit of Devils Tower roughly N corresponds to the original top surface of the 1340 extrusive body.

Asymmetry and Erosion of Devils Tower 75 vertical magm. f. 1490 (Halvorson, 1980) The asymmetry of Devils Tower, clearly vis- 1430 ible on the southeast-northwest proﬁ le (Fig. 1520 75 steep fabric 1C), could be easily explained in our structural with indicated plunge framework by the asymmetry of the original 1558 angle (Halvorson, 1980) extrusion (Fig. 16). For example, in two mod-

DT-1 ? 1460 els prepared with relatively diluted plaster sus- 1370 horizontal magm. f. 1 pension (D-19, D-21; Fig. 11), the analogue 4 0 ? (Halvorson, 1980) 0 magma rose within a tube-like conduit parallel to the walls of the model diatreme. While the 79 horizontal magm. f. with unknown lineation extrusive lobes on the right sides of both experimental bodies (Fig. 16; B side in Fig. 12D) 1340 plunge direction are conﬁ ned by the shallow angle of the maar crater walls, the opposite sides are molded 0100 m by a mound of popped-up diatreme filling that evacuated in front of advancing model Figure 15. Diagram showing the orientation of magmatic fabrics (magm. f.) on Devils Tower magma (Movie SI-2, experiment D-21 [see measured by Halvorson (1980) and by us (this study). Our study indicates only subhorizontal footnote 1]). The resulting contact between the fabrics of unknown plunge direction above elevation (contours) of 1490 m. Redrawn after extrusion and the mound could be character- Halvorson (1980). ized as rounded and as somewhat steeper than

Geosphere, April 2015 371

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/11/2/354/3335355/354.pdf by guest on 03 October 2021 Závada et al.

3 7 A B NW NE

UC UC LC LC J J T/P 4 T/P

2 5

1 6

NW NE

C D

T/P

Figure 16. Schematic block diagrams explaining the geological origin of Devils Tower and sketches of typical microstructures and mesoscale outcrops in the maar-diatremes volcanoes. (A) Phreatomagmatic eruptions excavate host rocks that are shattered and expelled together with fragments of phonolite magma within debris jets from the root zone (1) of a maar-diatreme volcano. Pilotaxitic porous texture of feldspar aggregates refl ects the rapid undercooling in this domain. Collapsing debris jets (3) form typical surge deposits (2) in the maar with variety of juvenile magmatic, sedimentary, and basement rock fragments and that can contain armored lapilli. The explosions excavate a cone-shaped diatreme with steep walls fi lled with phreatomagmatic deposits and a circular crater on the surface (maar) surrounded by a tuff ring (4), which is typical with matrix-dominated phreatomagmatic tuffs or agglomerates with large variety of clast sizes. Ejection of the debris from depth causes collapse (5) of the diatreme walls (marginal screen) and explains the bending of horizontal strata toward the tower. (UC—upper Colonnade, LC—lower colonnade, J—Jurassic, T/P—Triassic–Permian.) (B) The explosive activity shifts to a purely effusive phase of the volcano, when the ingress of groundwater necessary to trigger the phreatomagmatic explosions is insuffi cient or the lithostatic pressure at increasing depth is too high. The magma intruding the maar is extruded on the surface and forms a lava coulée (dome fl ow) fi lling the entire maar-diatreme crater. Faster convective cooling in the lower colonnade is provided by fl uids exsolved during crystallization of phonolite groundmass or released from the wet base of the maar (6). Associated microstructure sketch shows dilated trachytic texture of the phonolite with miarolitic cavities indicating horizontal stretching during the late stage of emplacement in the upper part of Devils Tower (7). The black lines indicate the trends of columnar jointing resolved from the two-dimensional thermal mathematical model. Dashed black lines on the front vertical section (B) indicate the isotherms. (C, D) Two perpendicular vertical cross sections across the lava coulée explain the possible role of the asymmetry of the original extrusion, also manifested by model D-21 (Fig. 11), on the resulting columnar jointing pattern. Note that in the section in D, where the extrusion had to overfl ow a mound of uplifted phreatomagmatic deposits, the cooling should produce curved columns (thin black lines in C and D) of the lower colonnade on the northwest side and straight columns on the southeast, where the extrusion was constrained by the fl at and shallowly inclined surface of the maar walls. Dashed red lines indicate the suture between the thermal fronts. The phreatomagmatic volcano and the lava coulée are eroded to leave behind only the central part of the solidifi ed lava coulée just above the feeding conduit, where the cooling produced the inverted-fan columnar jointing pattern, which is more resistant to erosion.

372 Geosphere, April 2015

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/11/2/354/3335355/354.pdf by guest on 03 October 2021 Devils Tower as a lava coulée?

on the opposite wall at the same level (Fig. Freise et al. (2003) constrain the minimum and Fink, 1993; Fink and Bridges, 1995) did 16). Thus in such a scenario, the curved col- temperature of extruding phonolite to be at not signiﬁ cantly hinder the growth and lateral umns of the northwest wall of Devils Tower least 850 °C. Phonolite magma of 850 °C with ﬂ ow of magma in the maar crater.

would form by cooling from the surface of the <1 wt% H2O corresponds to melt viscosities of uplifted mound, while the straight columns ~108–9 Pa·s (Whittington et al., 2001; Romano CONCLUSIONS of the southeast walls would form by cooling et al., 2003; Giordano et al., 2004). Incorpo- from the shallowly inclined and fl at bottom of rating the infl uence of phenocrysts (total 36 The evaluation of structural setting and the extrusive lobe emplaced along the walls of vol%) on melt viscosities of 108–9 Pa·s produces microstructural analysis of the magmatic and the maar crater. This hypothetical structural magma viscosities of ~4.5 × 108–9 Pa·s follow- phreatomagmatic rocks collected in the area analysis would further imply that: (1) the forming the equation of Dingwell et al. (2003) with suggests that emplacement of the Devils Tower ing extrusive fl ow advanced from southeast to infi nite viscosity at 0.6 phenocryst volume frac- phonolite monolith is associated with activity of northwest, fi nally forming an ~270-m-thick tion. Such viscosities are typical for the sug- a Cenozoic maar-diatreme volcano. The com- lava coulée; (2) most of the original extru- gested morphology of the coulée and/or low bined methods of analogue modeling simulat- sion was eroded from the northwest side of lava dome (Fink, 1980), although the extrusion ing magma intrusion into maar-diatremes and the remaining phonolite monolith forming the shape is also dictated by the yield strength and thermal mathematical modeling of cooling of tower; and (3) the diatreme should underlie the cooling rate of the magma and the topography magmatic bodies revealed that Devils Tower talus at the northwest side of the tower. The of the basal surface (Blake, 1990; Griffi ths and can be explained as a remnant of a low lava erosion of the tower advanced from the lateral Fink, 1997; Kerr et al., 2006). Both the vis- dome or coulée emplaced into the maar of a margins on the circumference of the extrusion cosity and yield strength of the magma likely maar-diatreme volcano. This new hypothesis is toward the interior by progressive collapse of further increased during fl ow at fi nal stages of illustrated by a perfect match of the columnar tilted subvertical columns that cleaved off the emplacement as the magma continuously crys- jointing pattern displayed on Devils Tower and remaining phonolite body primarily by frost tallized; this is recorded by the weak trachytic the thermal structure of the numerical model wedging (Tharp, 1987; Matsuoka, 2008). The texture of groundmass crystals. The transition (Fig. 14D). Although this hypothesis requires remaining monolith in its present form consists from pseudo-plastic to fi nal dilatant (strain- testing and further analysis and discussion, the of columns leaning against the central axis of thickening) rheology (Smith, 2000; Závada model explains the asymmetry of Devils Tower the tower and thus represents a relatively stable et al., 2011; Petford, 2009; Picard et al., 2011) and can be tested in the future by geophysical structure resistant to frost erosion. is evidenced by the miarolitic cavities in the methods such as combined gravity, magnetic, vicinity of large phenocrysts (Fig. 7A). Since and magnetotelluric surveys. Phonolite Magma Properties the aspect ratio of isothermal Bingham suspen- Our scenario is further corroborated by sion bodies is dictated by the yield strength another complex structural study of a similar The suggested extrusive form, a coulée or (Blake, 1990; Závada et al., 2009b), the origi- phonolite butte (Fig. 17A) in the Cenozoic vol- low lava dome, for the Devils Tower phonolite nal shape of the Devils Tower phonolite extru- canic province in central Europe (Závada et al., would be characterized by almost completely sion could be successfully reproduced by our 2011). Another analogous butte is found in the devolatilized and dehydrated magma. The experiments if we assume that the cooling rate French village of Ardéche in the Massif Central topology of dry solidus curve of Taylor and was rather slow with respect to emplacement (Roche de Borée) (Fig. 17B); however, no struc- MacKenzie (1975) for phonolite and experi- time of this relatively volumetric phonolite tural or petrological data are available for this mental modeling of phonolite composition by body, so that the cooling carapace (Griffi ths feature. Although our explanation for the geo-

A B

UC LC

Figure 17. Phonolite bodies similar to Devils Tower in Cenozoic volcanic provinces in Europe. (A) Phonolite body Bořeň (Czech Republic), view from the west, interpreted as a remnant of an extrusive lava dome emplaced into a maar-diatreme crater (Závada et al., 2011). The level of the contact between the colonnades of columns (UC—upper colonnade, LC—lower colonnade) is indicated by thick dashed line. (B) Roche de Borée in Ardéche, Massif Central (France), a phonolite body of similar shape and vertical columnar jointing in the upper part. Photographs by K. Mach.

Geosphere, April 2015 373

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/11/2/354/3335355/354.pdf by guest on 03 October 2021 Závada et al.

logical origin of Devils Tower is substantially DeGraff, J.M., and Aydin, A., 1993, Effect of thermal regime Halvorson, D.L., 1980, Geology and petrology of the Devils on growth increment and spacing of contraction joints Tower, Missouri Buttes, and Barlow Canyon area, different from previous published hypotheses, in basaltic lava: Journal of Geophysical Research, Crook County, Wyoming [Ph.D. thesis]: Grand Forks, we have presented an approach that gives clear v. 98, no. B4, p. 6411–6430, doi: 10 .1029 /92JB01709 . University of North Dakota, 163 p. guidelines for structural reconstruction of simi- DeWitt, E., Redden, J., Buscher, D., and Wilson, A.B., 1989, Jaeger, J.C., 1961, The cooling of irregularly shaped igneous Geologic map of the Black Hills area, South Dakota bodies: American Journal of Science, v. 259, p. 721– lar volcanic bodies. and Wyoming: U.S. Geological Survey Miscellaneous 734, doi: 10 .2475 /ajs .259 .10 .721 . Investigations Series Map I-1910, scale 1:250,000. Jaggar, T.A., 1901, Laccoliths of the Black Hills with a chap- ACKNOWLEDGMENTS Dingwell, D.B., Bagdassarov, N., Bussod, J., and Webb, S., ter by Ernest Howe on experiments illustrating intru- 2003, Magma rheology, in Luth, R. W., ed., Experi- sion and erosion: U.S. Geological Survey 21st Annual Permission for sample collection and fi eld work in ments at high pressure and applications to the Earth’s Report, p. 171–302. Devils Tower National Monument area from the U.S. mantle: Mineralogical Association of Canada Short Jančušková, Z., Schulmann, K., and Melka, R., 1992, Rela- Handbook SC21, p. 131–196. tion entre fabrique de la sanidine et mise en place des National Park Service (NPS; DETO-2009-SCI-0002) Duke, G.I., 2009, Black Hills–Alberta carbonatite–kimber- magmas trachytiques, exemple de massif de Hradište; is gratefully acknowledged. Permission for access in lite linear trend: Slab edge at depth?: Tectonophysics, Bohême du Nord: Geodinamica Acta, v. 5, p. 235–244. the study area was provided by the people of Lake v. 464, p. 186–194, doi: 10 .1016 /j .tecto .2008 .09 .034 . Jelínek, V., 1981, Characterization of magnetic fabric of Ranch at Missouri Buttes and the Driskill family Duke, G.I., Singer B.S., and DeWitt, E., 2002, 40Ar/39Ar rocks: Tectonophysics, v. 79, p. 63–67, doi: 10 .1016 at Devils Tower. We thank Jefferson Chang and laser incremental-heating ages of Devils Tower and /0040 -1951 (81)90110-4 . Stephen Holloway from the School of Geology and Paleocene-Eocene intrusions of the northern Black Kavanagh, J.L., Menand, T., and Sparks, R.S.J., 2006, An Geo physics, University of Oklahoma, and Mark Biel Hills, South Dakota and Wyoming: Geological Soci- experimental investigation of sill formation and propa- from the NPS for cooperation in the fi eld. This work ety of America Abstracts with Programs, v. 34, no. 6, gation in layered elastic media: Earth and Planetary p. 473. Science Letters, v. 245, p. 799–813, doi: 10 .1016 /j .epsl was funded in the frame of project ME09011 of the Duke, G.I., Carlson, R.W., Frost, C.D., Hearn B.C.J., and .2006 .03 .025 . program KONTAKT of the Ministry of Education Eby, G.N., 2014, Continent-scale linearity of kim- Kerr, R.C., Griffi ths, R.W., and Cashman, K.V., 2006, of the Czech Republic and supported by a Czech berlite-carbonatite magmatism, mid-continent North Formation of channelized lava fl ows on an uncon- Academy of Sciences Grant Agency Junior grant America: Earth and Planetary Science Letters, v. 403, fi ned slope: Journal of Geophysical Research, v. 111, (B301110703) for Prokop Závada. Prokop Závada p. 1–14, doi: 10 .1016 /j .epsl .2014 .06 .023 . B10206, doi: 10 .1029 /2005JB004225 . thanks Janek Hradilek for introducing him to the Dutton, C.E., and Schwartz, G., 1936, Notes on the jointing Kirchner, J.G., 1996, The geology of the Tomahawk Vol- landscape of the northern Great Plains. Jaroslav Lexa of the Devils Tower, Wyoming: Journal of Geology, canic Center, in Paterson, C.J., and Kirchner, J.G., eds., acknowledges support of VEGA Grant (Grant Agency v. 44, p. 717–728, doi: 10 .1086 /624472 . Guidebook to the geology of the Black Hills, South Effi nger, W., 1934, A report on the geology of the Devils Dakota: South Dakota School of Mines and Technol- of the Ministry of Education of the Slovak Republic Tower National Monument: Berkeley, California, ogy Bulletin, v. 19, p. 172–179. and Slovak Academy of Science) grant 162/11. Department of the Interior, National Park Service Field Kiver, E.P., and Harris, D.V., 1999, Geology of U.S. park- Division of Education, 14 p. lands (fi fth edition): New York, John Wiley & Sons, REFERENCES CITED Everson, D.D., and Roggenthen, W.M., 1988, Gravity and Inc., 912 p. magnetic models of the Tomahawk Diatreme, northern Kohl, T., and Hopkirk, R.J., 1995, Fracture—A simulation Arbaret, L., Diot, H., and Launeau, P., 1993, Le Suc phono- Black Hills, South Dakota, in Wyoming Geological code for forced fl uid-fl ow and transport in fractured, litique du Petit Gerbier (Velay, Massif Central); fab- Association 39th Annual Field Conference Guidebook, porous rock: Geothermics, v. 24, p. 333–343, doi:10 riques magnétique et magmatique: Comptes Rendus de Eastern Powder River Basin–Black Hills, p. 77–84. .1016 /0375 -6505 (95)00012 -F . l’Academie des Sciences, ser. 2, Mécanique, Physique, Fan, M., Quade, J., Dettman, D., and DeCelles, P.G., 2011, Konečný, V., and Lexa, J., 2003, Evolution of the phreato- Chimie, Sciences de l’Univers: Sciences de la Terre, Widespread basement erosion during the late Paleo- magmatic/extrusive/intrusive complex of the Bulhary v. 316, p. 1603–1610. cene–early Eocene in the Laramide Rocky Mountains maar-diatreme volcano in Southern Slovakia: Geolines Auer, A., Martin, U., and Németh, K., 2007, The Fekete- inferred from 87Sr/86Sr ratios of freshwater bivalve fos- (Praha), v. 15, p. 47–51. hegy (Balaton Highland Hungary) “soft-substrate” sils: Geological Society of America Bulletin, v. 123, Kratinová, Z., Závada, P., Hrouda, F., and Schulmann, K., and “hard-substrate” maar volcanoes in an aligned p. 2069–2082, doi: 10 .1130 /B30219 .1 . 2006, Non-scaled analogue modelling of AMS devel- volcanic complex—Implications for vent geometry, Fashbaugh, E., 1979, Geology of igneous extrusive and opment during viscous fl ow: A simulation on diapir- subsurface stratigraphy and the palaeoenvironmen- intrusive rocks in the Sundance area, Crook County like structures: Tectonophysics, v. 418, p. 51–61, doi: tal setting: Journal of Volcanology and Geothermal Wyoming [M.S. thesis]: Grand Forks, University of 10 .1016 /j .tecto .2005 .12 .013 . Research, v. 159, p. 225–245, doi:10 .1016 /j .jvolgeores North Dakota, 94 p. Lexa, O., Štípská, P., Schulmann, K., Baratoux, L., and .2006.06 .008 . Fink, J., 1980, Surface folding and viscosity of rhyolite Kröner, A., 2005, Contrasting textural record of two Blake, S., 1990, Viscoplastic models of lava domes, in Fink, fl ows: Geology, v. 8, p. 250–254, doi: 10 .1130 /0091 distinct metamorphic events of similar P-T conditions J.H., ed., Lava fl ows and domes: Emplacement mecha- -7613 (1980)8 <250: SFAVOR>2 .0 .CO;2 . and different durations: Journal of Metamorphic Geol- nisms and hazard implications: IAVCE Proceedings Fink, J.H., and Bridges, N.T., 1995, Effects of eruption ogy, v. 23, p. 649–666, doi: 10 .1111 /j .1525 -1314 .2005 in Volcanology Volume 2: Berlin, Springer-Verlag, history and cooling rate on lava dome growth: Bul- .00601 .x . p. 88–126. letin of Volcanology, v. 57, p. 229–239, doi:10 .1007 Lisenbee, A.L., and DeWitt, E., 1993, Laramide evolution Buisson, C., and Merle, O., 2002, Experiments on internal /BF00265423 . of the Black Hills uplift, in Snoke, A.W., et al., eds., strain in lava dome cross sections: Bulletin of Vol- Fischer, R., and Schmincke, H.-U., 1984, Pyroclastic rocks: Geology of Wyoming: Geological Survey of Wyoming canology, v. 64, p. 363–371, doi: 10 .1007 /s00445 -002 New York, Springer-Verlag, 472 p. Memoir 5, p. 374–412. -0213 -6 . Flores, R., and Ethridge, F., 1985, Evolution of intermontane Lisenbee, A.L., and Roggenthen, W., 1990, Diatremes and Cajz, V., Vokurka, K., Balogh, K., Lang, M., and Ulrych, J., fl uvial systems of Tertiary Powder River Basin, Mon- breccia pipes of the northern Black Hills, Wyoming, in 1999, The České středohoří Mts.: Volcanostratigraphy tana and Wyoming, in Flores, R.M., and Kaplan, S.S., Paterson, C.J., and Lisenbee, A.L., eds., Metallogeny and geochemistry: Geolines (Praha), v. 9, p. 21–28. eds., Cenozoic paleogeography of the west-central of gold in the Black Hills, South Dakota: Society of Campbell, M.E., Russell, J.K., and Porritt, L.A., 2013, Ther- United States: Rocky Mountain Paleogeography Sym- Economic Geologists Guidebook Series Volume 7, momechanical milling of accessory lithics in volcanic posium 3: Rocky Mountain Section, Society of Eco- p. 175–181. conduits: Earth and Planetary Science Letters, v. 377, nomic Paleontologists and Mineralogists, p. 107–126. Lorenz, V., 1973, On the formation of maars: Bulletin of Vol- p. 276–286, doi: 10 .1016 /j .epsl .2013 .07 .008 . Freise, M., Holtz, F., Koepke, J., Scoates, J., and Leyrit, H., canology, v. 37, p. 363–371. Carpenter, F.R., 1888, Notes on the geology of the Black 2003, Experimental constraints on the storage condi- Lorenz, V., 1986, On the growth of maars and diatremes Hills, in Carpenter, F.R., Preliminary report upon the tions of phonolites from the Kerguelen Archipelago: and its relevance to the formation of tuff-rings: Bul- geology, mineral resources and mills of the Black Contributions to Mineralogy and Petrology, v. 145, letin of Volcanology, v. 48, p. 265–274, doi:10 .1007 Hills of Dakota: Rapid City, Dakota Territory, Dakota p. 659–672, doi: 10 .1007 /s00410 -003 -0453 -2 . /BF01081755 . School of Mines Preliminary Report, p. 11–52. Giordano, D., Romano, C., Papale, P., and Dingwell, D.B., Lorenz, V., 2003, Maar-diatreme volcanoes, their formation, Cloos, H., and Cloos, E., 1927, Die Quellkuppe des Drachen- 2004, The viscosity of trachytes, and comparison with and their setting in hard-rock or soft-rock environ- fels am Rhein. Ihre Tektonik und Bildungsweise: basalts, phonolites, and rhyolites: Chemical Geology, ments: Geolines (Praha), v. 15, p. 72–83. Zeitschrift für Vulkanologie, v. 11, p. 33–40. v. 213, p. 49–61, doi: 10 .1016 /j .chemgeo .2004 .08 .032 . Lorenz, V., and Kurszlaukis, S., 2007, Root zone processes Corry, C.E., 1988, Laccoliths; mechanics of emplacement Griffi ths, R.W., and Fink, J.H., 1993, Effects of surface cool- in the phreatomagmatic pipe emplacement model and growth: Geological Society of America Special ing on the spreading of lava fl ows and domes: Journal and consequences for the evolution of maar-diatreme Paper 220, 110 p., doi:10 .1130 /SPE220 . of Fluid Mechanics, v. 252, p. 667–702, doi: 10 .1017 volcanoes: Journal of Volcanology and Geothermal DeGraff, J.M., and Aydin, A., 1987, Surface-morphology of /S0022112093003933. Research, v. 159, p. 4–32, doi:10 .1016 /j .jvolgeores columnar joints and its signifi cance to mechanics and Griffi ths, R.W., and Fink, J.H., 1997, Solidifying Bingham .2006 .06 .019 . direction of joint growth: Geological Society of Amer- extrusions: A model for the growth of silicic lava Martin, U., and Németh, K., 2004, Peperitic lava lake-fed ica Bulletin, v. 99, p. 605–617, doi: 10 .1130 /0016 -7606 domes: Journal of Fluid Mechanics, v. 347, p. 13–36, sills at Ság-hegy, western Hungary: A complex inter- (1987)99 <605: SMOCJA>2 .0 .CO;2 . doi: 10 .1017 /S0022112097006344 . action of a wet tephra ring and lava, in Breitkreuz, C.,

374 Geosphere, April 2015

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/11/2/354/3335355/354.pdf by guest on 03 October 2021 Devils Tower as a lava coulée?

and Petford, N., eds., Physical geology of subvolcanic Robinson, C., Mapel, J.W., and Bergendahl, M.H., 1964, Tȃmaș C.G., and Milesi, J.-P., 2002, Hydrovolcanic brec- systems: Geological Society of London Special Pub- Stratigraphy and structure of the northern and western cia pipe structures—General features and genetic cri- lication 234, p. 33–50, doi: 10 .1144 /GSL .SP .2004 .234 fl anks of the Black Hills uplift, Wyoming, Montana, teria. I. Phreatomagmatic breccias: Studia Universitatis .01 .04 . and South Dakota: U.S. Geological Survey Profes- Babeş-Bolyai Geologia, v. 47, p. 127–147. Martin, U., and Németh, K., 2005, Eruptive and deposi- sional Paper 404, scale 1:96,000, 134 p. Tarling, D., and Hrouda, F., 1993, The magnetic anisotropy tional history of a Pliocene tuff ring that developed in Romano, C., Giordano, D., Papale, P., Mincione, V., Ding- of rocks: New York, Chapman and Hall, 217 p. a fl uvio-lacustrine basin: Kissomlyó volcano (western well, D., and Rossi, M., 2003, The dry and hydrous vis- Taylor, D., and MacKenzie, W., 1975, A contribution to Hungary): Journal of Volcanology and Geothermal cosities of alkaline melts from Vesuvius and Phlegrean the pseudoleucite problem: Contributions to Miner- Research, v. 147, p. 342–356, doi: 10 .1016 /j .jvolgeores Fields: Chemical Geology, v. 202, p. 23–38, doi: 10 alogy and Petrology, v. 49, p. 321–333, doi: 10 .1007 .2005 .04 .019 . .1016 /S0009 -2541 (03)00208 -0 . /BF00376184 . Matsuoka, N., 2008, Frost weathering and rockwall erosion Ross, P.-S., and White, J.D.L., 2012, Quantifi cation of ves- Tharp, T.M., 1987, Conditions for crack propagation by in the eastern Swiss Alps: Long-term (1994–2006) icle characteristics in some diatreme-fi lling deposits, frost wedging: Geological Society of America Bulletin, observations: Geomorphology, v. 99, p. 353–368, doi: and the explosivity levels of magma-water interac- v. 99, p. 94–102, doi: 10 .1130 /0016 -7606 (1987)99 <94 10 .1016 /j .geomorph .2007 .11 .013 . tions within diatremes: Journal of Volcanology and :CFCPBF>2 .0 .CO;2 . Morrissey, M., Zimanowski, B., Wohletz, K., and Buettner, Geothermal Research, v. 245, p. 55–67, doi: 10 .1016 /j Tikoff, B., and Maxson, J., 2001, Lithospheric buckling of R., 2000, Phreatomagmatic fragmentation, in Sigurds- .jvolgeores .2012 .07 .006 . the Laramide foreland during Late Cretaceous and son, H., ed., Encyclopedia of volcanoes: San Diego, Russell, I.C., 1896, Igneous intrusions in the neighborhood Paleogene, western United States: Rocky Mountain California, Academic Press, p. 431–445. of the Black Hills of Dakota: Journal of Geology, v. 4, Geology, v. 36, p. 13–35, doi: 10 .2113 /gsrocky .36 Nagata, T., 1961, Rock magnetism: Tokyo, Maruzen, 366 p. p. 23–43, doi: 10 .1086 /607420 . .1 .13 . Paterson, S.R., Fowler, T.K., Schmidt, K.L., Yoshinobu, Scarpati, C., Cole, P., and Perrotta, A., 1993, The Neapoli- Valentine, G.A., and White, J., 2012, Revised conceptual A.S., Yuan, E.S., and Miller, R.B., 1998, Interpret- tan Yellow Tuff—A large-volume multiphase eruption model for maar-diatremes: Subsurface processes, ener- ing magmatic fabric patterns in plutons: Lithos, v. 44, from Campi Flegrei, southern Italy: Bulletin of Vol- getics, and eruptive products: Geology, v. 40, p. 1111– p. 53–82, doi: 10 .1016 /S0024 -4937 (98)00022 -X . canology, v. 55, p. 343–356, doi: 10 .1007 /BF00301145 . 1114, doi: 10 .1130 /G33411 .1 . Petford, N., 2009, Which effective viscosity?: Mineralogi- Schulz, R., Buness, H., Gabriel, G., Pucher, R., Rolf, C., Valentine, G.A., Graettinger, A.H., and Sonder, I., 2014, cal Magazine, v. 73, p. 167–191, doi: 10 .1180 /minmag Wiederhold, H., and Wonik, T., 2005, Detailed inves- Explosion depths for phreatomagmatic eruptions: Geo- .2009 .073 .2 .167 . tigation of preserved maar structures by combined physical Research Letters, v. 41, p. 3045–3051, doi:10 Picard, D., Arbaret, L., Pichavant, M., Champallier, R., geophysical surveys: Bulletin of Volcanology, v. 68, .1002 /2014GL060096 . and Launeau, P., 2011, Rheology and microstructure p. 95–106. Varet, J., 1971, Structure et mise en place des massifs of experimentally deformed plagioclase suspensions: Sigurdsson, H., ed., 2000, Encyclopedia of volcanoes: San phonolitique du Cantal (Auvergne, France): Geolo- Geology, v. 39, p. 747–750, doi: 10 .1130 /G32217 .1 . Diego, California, Academic Press, 1417 p. gische Rundschau, v. 60, p. 948–970, doi: 10 .1007 Pirsson, L., 1894, Description of the character of the igneous Smith, J.V., 2000, Textural evidence for dilatant, shear /BF02046530 . rocks making up Mato Teepee and the Little Missouri thickening) rheology of magma at high crystal con- White, J., and Ross, P.-S., 2011, Maar-diatreme volcanoes: Buttes: American Journal of Science, v. 47, p. 341– centrations: Journal of Volcanology and Geothermal A review: Journal of Volcanology and Geothermal 346, doi:10 .2475 /ajs .s3 -47 .281 .341 . Research, v. 99, p. 1–7, doi:10 .1016 /S0377 -0273 Research, v. 201, p. 1–29, doi: 10 .1016 /j .jvolgeores Pollard, D., and Johnson, A., 1973, Mechanics of growth of (99)00191 -2 . .2011 .01 .010 . some laccolithic intrusions in the Henry Mountains, Smith, J.V., Miyake, Y., and Oikawa, T., 2001, Interpretation Whittington, A., Richet, P., Linard, Y., and Holtz, F., 2001, Utah, II. Bending and failure of overburden layers and of porosity in dacite lava domes as ductile-brittle fail- The viscosity of hydrous phonolites and trachytes: sill formation: Tectonophysics, v. 18, p. 311–354, doi: ure textures: Journal of Volcanology and Geothermal Chemical Geology, v. 174, p. 209–223, doi: 10 .1016 10 .1016 /0040 -1951 (73)90051-6 . Research, v. 112, p. 25–35, doi: 10 .1016 /S0377 -0273 /S0009 -2541 (00)00317 -X . Price, N., and Cosgrove, J.W., 1990, Analysis of geological (01)00232 -3 . Závada, P., Schulmann, K., Lexa, O., Hrouda, F., Haloda, J., structures: Cambridge, Cambridge University Press, Smith, M.E., Carroll, A.R., and Mueller, E.R., 2008, Ele- and Týcová, P., 2009a, The mechanism of fl ow and fab- 502 p. vated weathering rates in the Rocky Mountains during ric development in mechanically anisotropic trachyte Rakovan, J., 2006, Diatreme: Rocks and Minerals, v. 81, the Early Eocene Climatic Optimum: Nature Geosci- lava: Journal of Structural Geology, v. 31, p. 1295– p. 153–154, doi: 10 .3200 /RMIN.81 .2 .153 -154 . ence, v. 1, p. 370–374, doi: 10 .1038 /ngeo205 . 1307, doi: 10 .1016 /j .jsg .2009 .04 .002 . Redden, J., Obradovich, J., Naeser, C., Zartman, R., and Sparks, R.S.J., Murphy, M.D., Lejeune, A.M., Watts, R.B., Závada, P., Kratinová, Z., Kusbach, V., and Schulmann, K., Norton, J., 1983, Early Tertiary age of pitchstone in the Barclay, J., and Young, S.R., 2000, Control on the 2009b, Internal fabric development in complex lava northern Black Hills, South Dakota: Science, v. 220, emplacement of the andesite lava dome of the Soufriere domes: Tectonophysics, v. 466, p. 101–113, doi: 10 p. 1153–1154, doi: 10 .1126 /science.220 .4602 .1153 . Hills Volcano, Montserrat by degassing-induced crys- .1016 /j .tecto .2008 .07 .005 . Reynolds, D.L., 1954, Fluidization as a geological process, tallization: Terra Nova, v. 12, p. 14–20, doi:10 .1046 /j Závada, P., Dědeček, P., Mach, K., Lexa, O., and Potužák, and its bearing on the problem of intrusive granites: .1365 -3121 .2000 .00267 .x . M., 2011, Emplacement dynamics of phonolite magma American Journal of Science, v. 252, p. 577–613, doi: Spry, A., 1962, The origin of columnar jointing, particularly into maar-diatreme structures—Correlation of fi eld, 10 .2475 /ajs .252 .10 .577 . in basalt fl ows: Australian Journal of Earth Sciences, thermal modeling and AMS analogue modeling data: Robinson, C., 1956, Geology of Devils Tower National v. 8, p. 191–216, doi: 10 .1080 /14400956285270061 . Journal of Volcanology and Geothermal Research, Monument, Wyoming: U.S. Geological Survey Bul- Sutherland, W., 2008, Geologic map of the Devils Tower v. 201, p. 210–226, doi: 10 .1016 /j .jvolgeores .2010 .07 letin 1021–1, 13 p. 30′ × 60′ Quadrangle, Crook County, Wyoming, Butte .012 . Robinson, C., and Davis, R.E., 1995, Geology of Devils and Lawrence Counties, South Dakota, and Carter Ziegler, P.A., 1994, Cenozoic rift system of western and Tower National Monument: Devils Tower, Wyoming, County, Montana: Wyoming State Geological Survey central Europe—An overview: Geologie en Mijnbouw, Devils Tower Natural History Association, 96 p. Map Series MS-81, scale 1:100,000. v. 73, p. 99–127.

Geosphere, April 2015 375

Downloaded from http://pubs.geoscienceworld.org/gsa/geosphere/article-pdf/11/2/354/3335355/354.pdf by guest on 03 October 2021