University of Portland Pilot Scholars Environmental Studies Faculty Publications and Environmental Studies Presentations

10-10-1986 North American Jurassic Apparent Polar Wander: Implications for Plate Motion, Paleogeography, and Cordilleran Tectonics Steven R. May

Robert F. Butler University of Portland, [email protected]

Follow this and additional works at: http://pilotscholars.up.edu/env_facpubs Part of the Environmental Sciences Commons, and the Geophysics and Seismology Commons

Citation: Pilot Scholars Version (Modified MLA Style) May, Steven R. and Butler, Robert F., "North American Jurassic Apparent Polar Wander: Implications for Plate Motion, Paleogeography, and Cordilleran Tectonics" (1986). Environmental Studies Faculty Publications and Presentations. 10. http://pilotscholars.up.edu/env_facpubs/10

This Journal Article is brought to you for free and open access by the Environmental Studies at Pilot Scholars. It has been accepted for inclusion in Environmental Studies Faculty Publications and Presentations by an authorized administrator of Pilot Scholars. For more information, please contact [email protected]. JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 91, NO. B11, PAGES 11 1 519-11,544, OCTOBER 10, 1986

NORTH AMERICAN JURASSIC APPARENT POLAR WANDER1 IMPLICATIONS FOR PLATE MOTION, PALEOGEOGRAPHY AND CORDILLERAN TECTONICS

Steven R. Mayl and Robert F. Butler

Department of Geosciences, University of Arizona, Tucson

Abstract. Eight paleomagnetic poles are con­ Paleomagnetic studies for cratonic North sidered to be reliable Jurassic reference poles America have been overshadowed in recent years by for cratonic North America. These poles form a the popularity of using paleomagnetism to con­ consistent chronological progression defining two strain the motion histories of suspect terranes arcuate tracks of apparent polar wander (APW) within the western Cordillera [Beck, 1976, 19801 from Sinemurian through Tithonian time (203-145 Hillhouse, 19771 Hillhouse and Gromme, 19801 Ma). Combined with reliable Triassic and Creta­ Irving et al., 1985]. Unfortunately, our under­ ceous reference poles, the resulting path is well standing of the North American APW path is not so modeled by paleomagnetic Euler pole (PEP) analy­ advanced as to warrant this neglect especially sis and is significantly different from previous for certain time intervals like the Jurassic. APW compilations. These differences reflect Reliable estimates of relative latitudinal dis­ differences in original data sets, modes of placements of suspect terranes are ultimately analysis, and geologic time scales and translate constrained by the accuracy of cratonic reference into substantial and important differences in poles, yet perusal of recent compilations auch as paleolatitude estimates for cratonic North Irving and Irving [ 1982] and Harrison and Lindh America. PEP analysis reveals two cusps, or (1982] reveals intervals of geologic time for changes in the direction of APW1 one in the Late which confidence parameters associated with Triassic to Early Jurassic (Jl) and one in the reference poles are very large. Such uncertain­ Late Jurassic (J2). The Jl cusp represents the ties translate directly into imprecise and change in North American absolute plate motion potentially inaccurate estimations of the paleo­ aBSociated with rifting of the central Atlantic la ti tud ina l history of North America. The and Gulf of Mexico, while the J2 cusp correlates Jurassic has been a particularly blatant example temporally with the marine magnetic anomaly M21 of this problem because of the paucity of well plate reorganization and to various North dated, reliable paleopoles, coupled with an American intraplate tectonomagmatic events (e.g., unusually large amount of apparent polar wander. Nevadan Orogeny). Analysis of pole progression The Late Triassic-Jurassic North American APW along the Jl to J2 and J2 to Cretaceous APW path records the opening and early plate motion tracks indicates constant angular plate velocity evolution of the central Atlantic Ocean basin of 0.6°-o. 1° /m.y. from 203 to 150 Ma followed by [Steiner, 1975, 1983]. Geometric analysis of the significantly higher velocity from 150 to 130? path can be directly related to plate reorganiza­ Ma. Late Triassic-Jurassic reference poles indi­ tion and North American absolute motion within cate more southerly paleolatitudes for cratonic the geologically-geophysically constrained rift North America than have previous compilations and drift history. The timing of first-order requiring modification of displacement scenarios changes in the shape of the APW path as deduced for suspect terranes along the western Cordil­ by paleomagnetic Euler pole (PEP) analysis lera. [Gordon et al., 1984] corresponds remarkably well with various global and regional plate and intra­ Introduction plate tectonic events suggesting causal relation­ ships. The time scale used is that of Harland et An apparent polar wander (APW) path is a time al. (1982]. sequence of paleomagnetic poles that records the paleolatitude and azimuthal orientation of a North American Jurassic APW plate within the dipolar geomagnetic field [Irving, 1977, 1979]. Invocation of the axial Historical Development of the JuraBBic APW Path geocentric dipole hypothesis permits sequential palaeogeographies to be constructed within a The first Jurassic paleomagnetic results from reference frame tied to the rotation axis. APW North America were described by Collinson and paths contain information regarding both the Runcorn (1960]. Poles from the Kayenta and direction and velocity of plate motion and there­ Carmel formations on the Colorado Plateau were fore are fundamental to analyses of plate kine­ used to construct the APW path shown in Figure matics, terrane displacements, and paleogeog­ la. As was common during this time, the calcu­ raphy. For this reason, APW paths require lated poles were based solely on directions of constant revision and reinterpretation as new natural remanent magnetism (NRM) without aid of paleomagnetic, geochronologic, and tectonic data current demagnetization techniques. In the case become available. of the Kayenta and Carmel Formations, NRM data provided very poor estimates of true pole posi­ t ions. Subsequent work has shown that these 1 Now at Exxon Production Research Company, units have significant Cenozoic or present field Houston, Texas. secondary overprints which have been successfully removed from Kayenta samples but not from the Copyright 1986 by the American Geophysical Union. Carmel Formation [Steiner and Helsley, 19741 Steiner, 1983]. The Mesozoic APW path con­ Paper number 6B5934. structed by Collinson and Runcorn [1960] shows a 0148-0227/86/006B-5934$05.00 gle track of polar motion from Early Triassic 11 ,519 11,520 May and Butler1 Jurassic Apparent Polar Wander

Fig. 1. ''Late Triaaaic" through Juraaaic North American APW paths. (a) Collinson and Runcorn [1960); Trc, Triassic Chugwater Formation; Trm, Triassic Moenkopi Formation; Trn, "Triassic" Newark Group rocks; Jk, Jurassic Kayenta Formation; Jc, Jurassic Carmel Formation. (b) Irving and Park [1972); TR, Triassic; J, Jurassic; and K, Cretaceous mean poles with associated A95 confidence circles. (c) Harrison and Lindh [1982). (d) Irving and Irving [1982), Figures le and ld were constructed with a "sliding-window" technique and show mean pole locations with A95 confidence circles. Mean ages of selected reference poles are shown in millions of years.

Chugwater and Moenkopi Formation poles through a was baaed on an average of poles from the White "Late Triassic Newark Formation" pole to the Mountain Magma Series [Opdyke and Wenaink, 1966), Jurassic Carmel and Kayenta poles, the average of the Anticoati Island diabase dike [Larochelle, which was indistinguishable from the geographic 1971), the Island Intrusions [Symons, 1970), and north pole. the Kayenta Formation pole of Collinson and Irving [1964) concluded that there were no Runcorn [ 1960]. reliable Jurassic paleopolea for North America Ironically, we now recognize that much of the and pointed out that the Kayenta and Carmel early paleomagnetic work by DuBois et al. [1957), Formation results of Collinson and Runcorn [1960) Opdyke [1961), deBoer [1967, 1968), and Beck were biased by present field overprint and were [1972) on Newark Supergroup and related rocks of not representative of the Jurassic paleofield. the northeastern United States was applicable to However, Irving and Park [1972) published a mean Jurassic APW. However, until the late 1970s, Juraaaic pole which, like the earlier Collinson these rocks were considered to be Late Triassic and Runcorn result, was statistically coincident rather than Early Jurassic. with the geographic pole (Figure lb). Thia pole McElhinny [1973) included only two paleopoles May and Butler1 Jurassic Apparent Polar Wander 11 '521 within hia Jurassic mean pole, those being the appear to be any satisfactory scheme free of White Mountain Magma Series pole [Opdyke and subjectivity. Although basically employing a 30- Wenaink, 1966] and the "Appalachian Mesozoic m.y. sliding window, Harrison and Lindh [1982] dikes" pole of deBoer [1967]. The resultant mean diacuaa a modification for constructing APW paths Jurassic pole at 76°N, 142°E lacked an associated baaed on weighting individual paleopoles ac­ confidence oval but waa used to define a track of cording to their "information content." Part of APW connecting Triassic and Cretaceous poles this technique involves weighting poles depending exclusive of the north pole. on the amount of overlap between the age range At about this time, a second generation of associated with the pole and the window being ~~ozoic paleomagnetic data from Triassic and calculated. Harrison and Lindh [1982] show that Jurassic sediments on the Colorado Plateau and age weighting and other somewhat more subjective from the eastern United States became available. weighting parameters can cause significant dif­ The work of Steiner and Helsley [1972, 1974, ferences between alternative APW paths especially 1975], Smith [1976], Steiner [1978], and Smith during intervals with low pole density, rapid and Noltimier [1979] greatly improved our under­ APW, and poor age control. The Jurassic interval standing of Jurassic APW and demonstrated of North American APW has suffered from all of unquestionably that the path did not pass through these problems. The most important conclusion of the geographic pole but tracked from Triassic to Harrison and Lindh [1982] ia that the fundamental Cretaceous poles along a band of latitude between factor producing variation in APW paths ia selec­ 60° and 70°N (present coordinates) [Steiner, tion of a~ original data base. 1975]. The second generation of APW paths Gordon et al. [1984] have suggested that APW [Irving, 1977, 1979; Van Alstine, 1979; Harrison paths can be generated by "paleomagnetic Euler and Lindh, 19821 Irving and Irving, 1982] have pole" (PEf) analysis. Their methodology assumes more or less approximated this latitudinal track that APW paths are composed of small circle seg­ of APW (Figures le and ld). However, many of ments and that deviation of any pole from a best these compilations include less reliable paleo­ fit small circle reflects inherent inaccuracy of polea which tend to bias average reference poles paleomagnetic techniques not apparent polar toward high latitudes. wander. Upon calculating the best fit small circle approximation to a track of AfW, Gordon et Constructing APW Paths1 Technigues and Critiguea al. [1984] collapse the data onto a line describing constant angular pole displacement as As more paleomagnetic data became available a function of age. Moving back into pole space, for all the major continents, the accepted pro­ they convert the PEP-APW model into a series of cedure for calculating APW paths changed. From time incremented "reference poles" whose geometry 1956 to 1977, the standard technique waa to group and age progression are constrained by the all paleomagnetic poles according to geologic original set of paleopoles but whose actual posi­ period and calculate mean reference poles of tions need not correspond to any of the original period duration. Van Alatine and deBoer [1978] data. suggested a technique for constructing APW paths PEP analysis is a very useful tool for that included demarcation of equal time intervals modeling APW and associated plate motion, but within which poles would be averaged. They synthetic reference poles thus derived should not pointed out that using geologic periods is be used as the sole basis for constraining abao­ unattractive because such periods are both long lu te paleogeographiea. PEP reference poles are and of unequal duration. This tends to decrease baaed on a forced fit of paleopole data to a the precision and usefulness of APW paths because plate tectonic model. Any inaccuracy in the details are overly smoothed and because rates of model will generate inaccurate reference poles APW cannot be readily estimated by the relative and information inherent in the original data separation between reference poles. will be lost. Reference poles for paleotectonic­ Irving [1977] also used a nonperiod standard pa leogeograph ic calculations should be baaed on time window averaging technique to generate poat­ actual paleomagnetic data and not on synthetic, Devonian reference poles for North America. Un- model dependent approximations. A detailed and 1 ike Van Alstine and deBoer'a 22-m.y. window, defensible application of paleomagnetic data Irving used a sliding window average of 40-m.y. requires calculation of the moat appropriate duration that waa incremented at 10 m.y. steps. reference pole compatible with the age of the Irving [1979] and Irving and Irving [1982] have desired reconstruction. subsequently used the same technique but with a Although age weighting may be useful in con­ 30-m.y. duration window. Although useful for junction with the sliding-window technique of APW illustrating the first-order changes in APW, the path construction, weighting schemes in general sliding window technique maaka some of the are usually subjective and do not necessarily detailed structure present in the raw paleopole yield increased accuracy. The moat important data aet. Hairpins or cusps (sharp changes in factor controlling the accuracy of, and varia­ the direction of APW) are heavily smoothed, and bility between, APW paths is the reliability of boundaries between episodes of rapid and slow APW the selected data base. Use of the terms become blurred. Also, previous Jurassic refer­ "reliable" and "unreliable" in this paper ence poles generated with the sliding window reflects our judgement as to whether or not a technique have been strongly biased into inac­ particular study meets the designated minimum curately high latitudes by the inclusion of acceptance criteria. These criteria include several unreliable high-latitude poles. demagnetization behavior, the number of sites, Various weighting schemes have been diacuaaed Fisher statistical parameters of the data aet, recently by Harrison and Lindh [1982] and Gordon age uncertainty, and geologic setting and are et al. [1984]. Unfortunately, there does not discussed in detail in the appendix. "Unreli- 11,522 May and Butler1 Jurassic Apparent Polar Wander

TABLE 1. North American Jurassic Reference Poles.

Symbol Pole Pole * on Age, Latitude Longitude A9S• Reference Pole Figures Age Ma ON OE deg.

Upper Morrison UM late Tithonian 145 67.6 161.9 3.9 1 Formation Lower Morrison LM early Tithonian 149 61.4 142.3 4.2 1 Formation Glance G Rb/Sr 151+2 62.7 131.5 6.3 2 Conglomerate Corral Canyon cc Rb/Sr 172:!:5.8 61.8 116.0 6.2 3 Rocks Newark Trend NTII Ar/Ar 179+3 65.3 103.2 1.4 4 Group II Newark Trend NTI Ar/Ar 195+4 63.0 83.2 2.3 4 Group I Ka yen ta K Pliensbachian 194-200 62.1 70.2 6.3 5 Formation Wingate w Sinemurian 200-206 59.0 63.0 8.0 6 Formation

*95l confidence angle about pole. References1 1, Steiner and Helsley [1975); 2, Kluth et al. [1982); 3, May et al. [this issue]; 4, Smith and Noltimier [1979); 5, Steiner and Helsley [1975); 6, Reeve [1975) from Gordon et al. [1984). able" does not neceaaarily mean that the original rocks and both were included in an average 200 Ma science was "wrong" or "sloppy" but commonly reference pole. Using the preferred time scale reflects complexity or uncertainty associated of Harland et al. [1982), the beat pick for the with critical parameters of a paleomagnetic data absolute age of the Chinle is 220-230 Ma or set which detract from its usefulness as a approximately 25-30 m.y. older than the Early reference pole. Jurassic (Pliensbachian) Newark trend rocks. Our philosophy in constructing an APW path has Such examples are common within compilations of been to select only high-quality paleomagnetic Late Triassic and Jurassic reference poles for poles and to evaluate the time sequence of these North America and have led to imprecision and original data. We acknowledge the contribution inaccuracy. Although the Harland et al. [1982] which has been made by the sliding-window-type time scale will almost certainly experience analysis but object to the smoothing of useful revision, the consistency observed between pre­ details which will result. Smoothing techniques dicted ages, radiometric ages, and relative pole can be effective at filtering random errors but positions is encouraging. The primary conclu­ will reinforce the unwanted bias of systematic sions of this analysis would be the same had we errors. In the case of Jurassic APW, paleopoles used the Decade of North American Geology (DNAG) have commonly been polluted by unremoved late time scale [Palmer, 1983], Mesozoic, Cenozoic, and present field overprints. As our knowledge of both APW and geologic time Inclusion of such poles has resulted in inac­ scales becomes more sophisticated, it is impor­ curately high latitudes for reference poles tant that any time sequence analyses be accom­ generated through sliding-window techniques. panied by a statement of the time scale used. Furthermore, as interdisciplinary synthesis of Geologic Time Scales global and regional tectonics depends largely on recognizing temporal coincidence, it is impera­ Because APW paths are commonly based on a data tive that all relevant data be analyzed with the set including paleopoles from paleontologically same time scale. As discussed later, part of the dated sedimentary rocks as well as radiometrical­ apparent paleomagnetic discordancy for certain ly dated igneous rocks, the choice of a geologic Cordilleran terranes can be traced directly to time scale can influence spatiotemporal inter­ inaccurate reference poles generated with "out­ pretations. Time scales are especially critical dated" geologic time scales. to the interpretation of Jurassic APW because the "absolute" ages associated with Jurassic period, Jurassic Paleomagnetic Poles epoch, and age boundaries have undergone signifi­ cant revision in the past 20 years. There are eight reliable paleomagnetic poles For example, the Chinle Formation of Late from Jurassic age rocks of cratonic North America Triassic (Carnian-Norian) age was assigned an (Table 1). These include Early Jurassic poles absolute age of 199 Ma by Irving and Irving from the Wingate Formation [Reeve, 1975), the [1982) using a time scale similar to Van Eyainga Kayenta Formation [Steiner and Helsley, 1974], [1975]. The Chinle pole was therefore considered and the Newark Trend Group I intrusive rocks nearly correlative with a 195:!:5 Ma radiometrical­ [Smith and Noltimier, 1979); Middle Jurassic ly calibrated pole from the Newark Trend igneous poles from the Newark Trend Group II intrusive May and Butlers Jurassic Apparent Polar Wander 11 '523

Fig. 2. Revised Triassic-Early Cretaceous North American APW path. (a) Stereographic north polar projection showing reliable reference poles as listed in Tables 1 and 2. Symbols for Jurassic poles are W, Wingate Formation; K, Kayenta Formation; NTI, Newark Trend Group I; NTII, Newark Trend Group II; CC, Corral Canyon; G, Glance Conglomerate (Canelo Hilla); LM, lower Morrison Formation; UM, upper Morrison Formation. Other poles include RP, Red Peak Formation of Chugwater Group (two poles); SB, State Bridge Formation; M, Moenkopi Formation; MI, Manicouagan Impact Structure; C, Chinle Forma­ tion; and KA, Cretaceous average pole of Mankinen [1978). Mean pole locations are shown by solid circles and associated A95 confidence regions. (b) Same as Figure 2a, but with poles from the Colorado Plateau corrected for 3.8° clockwise rotation as in Table 3. Modified poles include UM, LM, K, W, and M. rocks [Smith and Noltimier, 1979), and the Corral figures and Table 2 is an average Early-Middle Canyon sequence [May et al., this issue]; and Cretaceous pole calculated by Mankinen [1978). Late Jurassic poles from the Glance Conglomerate Although certain of the eight poles used in this [Kluth et al., 1982) and the lower and upper average are of questionable reliability, their Morrison Formation [Steiner and Helsley, 1975). consistency and tight clustering suggest that the The epatiotemporal distribution of these poles mean pole is a good approximation of the paleo­ defines a consistent eastward progression of APW field during the Cretaceous stilletand from Sinemurian to late Tithonian time (Figure ("etilletand" as used here refers to an interval 2). Detailed discussions of each of these poles of essentially no APW). The oldest of the reli­ as well as poles used in previous analyses but able poles in this group is the Monteregian Hille here considered unreliable are presented in the intrusive pole which has been assigned a mean appendix. We wish to emphasize that our purpose K/Ar age of 126+6 Ma, while the youngest pole is in relegating these discussions to an appendix is from the Niobrara Formation at approximately 85- to facilitate the coherent flow of our main ideas 90 Ma [Shive and Frerichs, 1974). Recently and conclusions and not to deemphasize the impor­ reported fission track and Rb/Sr dates from the tance of that information. To the contrary, the Monteregian Hille intrusives show two clusters of appendix represents the basis of our analysis and ages at about 118 Ma and 136 Ma [Eby, 1984). contains numerous discussions of data interpreta­ Paleomagnetic results from this intrusive series tion which, to varying degrees, guide our conclu­ need reevaluation, but the new geochronology sions. suggests that the Cretaceous standstill may have begun as early as 136 Ma. The lack of signifi­ Triassic and Cretaceous Polee cant APW during the interval from ? 130 to 85 Ma justifies our use of a single mean pole in later Similar analysis of Triassic and Early Creta­ PEP analysis of the Late Jurassic APW track. It ceous paleomagnetic poles provides a context of is important to realize that the episode of rapid reliable pre- and poet-Jurassic APW. These poles Late Jurassic APW ended by at least 126+6 Ma, and are not discussed in detail but are listed in probably somewhat earlier, although there are no Table 2 and shown in Figure 2. Early Triassic reliable poles of certain Berriasian age. poles from the Chugwater Group and the Moenkopi and State Bridge formations illustrate the trend ! Revised Jurassic ~ Path of the pre-Jurassic APW track, although the only reliable Late Triassic poles for North America Using the time scale of Harland et al. [1982), are the Chinle Formation pole [Reeve and Helsley, Jurassic paleopolee form a consistent chronologic 1972) of Carnian-Norian age and the Manicouagan progression that defines a path of APW from the pole [Robertson, 1967; Larochelle and Currie, Sinemurian through the Tithonian (203-145 Ma). 1967). Combined with reliable Late Triassic and Early The single Cretaceous pole shown in various Cretaceous paleopolee, the resulting APW path is 11,524 May and Butlers Jurassic Apparent Polar Wander

TABLE 2. Triassic and Cretaceous North American Reference Poles.

Symbol Age Pole Pole on Age, Latitude Longitude A95• Referencet Pole Figures Ma ON OE deg.

Cretaceous KA mean 130-85 68.0 186.0 2.2 1 Average Manicouagan MI K/Ar 215±5 58.8 89.9 5.8 2 Structure Chin le c Carnian-Norian 220-230 57.7 79.1 7.0 3 Formation Moenkopi M Early-Middle 231-248 57.0 100.3 5.3 4 Formation Triassic State Bridge SB Early TriaBBic 243-248 52.0 107.0 3.0 5 Formation Red Peak RP Early TriaBBic 243-248 46.6 113.5 1.9 6 Formation Red Peak RP Early TriaBBic 243-248 45.4 115.3 4.1 7 Formation

See footnote for Table 1. tReferences: 1, Mankinen [1978]1 2, Robertson [1967] and Larochelle and Currie [1967]1 3, Reeve and Helsley [1972]1 4, Baag and Helsley [1974]1 5, Christensen [1974] from Gordon et al. [1984]; 6, Shive et al. [1984]; 7, Herrero-Brevera and Helsley [1983]. different from previously published compilations and Jurassic reference poles in Irving and of Irving [1977], Van Alstine and deBoer [1978], Irvings' compilation predict higher paleolati­ Briden et al. [1981], Irving and Irving [1982], tudes for North America than does our revised and Harrison and Lindh [1982]. These differences APW path. The magnitude of this difference include a marked cusp in the Early Jurassic, as translates into differences in predicted mean also recognized by Gordon et al. [1984], rela­ paleolatitude at San Francisco of approximately tively low latitudes for Late Triassic through 750 km at 200 Ma and 600 km at 170 Ma. Such Late Jurassic reference poles (58°-63°N present differences obviously affect interpretations con­ coordinates), and a second cusp in the Late cerning the allochthoneity of Cordilleran suspect Jurassic. Each of these features has implica­ terranes. tions for North American plate motion and paleo­ In conjunction with differences in the abso­ latitudes and for Cordilleran paleogeography. lute positions of reference poles, the basic To illustrate the important characteristics of geometries of the paths are dissimilar. The the revised Jurassic APW path, we compare it to moving average technique of Irving and Irving the recent and popular path of Irving and Irving (and others) has the effect of smoothing changes [1982) (Figure ld). It is important to remember in direction of APW. Because abrupt changes in that these paths were constructed in fundamental­ APW may be correlated with important plate ly different ways and with quite different reorganizations and intraplate tectonic events geological time scales. Our technique is to [e.g., Beck, 1984), this is an important dif­ generate an APW path simply as a time sequence of ference. Numerous correlations can be hypothe­ high-quality paleopoles. This provides access to sized between the structure recognized in the the maximum amount of information inherent in the revised APW path and North American tectonics. raw data. Irving and Irving [1982], on the other hand, use a sliding-window averaging technique Paleopoles From the Colorado Plateau with less rigorous data selection to reveal first-order patterns of APW. Until recently, much of our knowledge of As discussed previously, differences in geo­ Jurassic APW was based on paleomagnetic studies logic time scales can profoundly influence the from sedimentary rocks on the Colorado Plateau interpretation of APW especially when a sliding­ [Steiner, 1983]. Plateau-derived poles still window averaging technique is used. Although not comprise 501 of the present list of reliable data cited, the time scale used by Irving and Irving and are critical for understanding Early and Late [1982) was similar to that of Van Eysinga [1975] Jurassic features of the APW path. Recently, which places the Triassic-Jurassic boundary at there has been some discussion that the Colorado approximately 195 Ma and the Jurassic-Cretaceous Plateau may have experienced a small clockwise boundary at 141 Ma. The 200 Ma reference pole rotation with respect to the rest of the craton was therefore constructed as a Late Triassic pole in post-Jurassic time. Before proceeding with a at 63°N, 92°E, A95-4°. The correlative pole in discussion of Jurassic APW analysis, we must terms of absolute age in our revised path is the first address the question of tectonic rotation Sinemurian (200-206 Ma) Wingate Formation pole of the Colorado Plateau. located at 59°N, 63°E, A95•8°. Depending on On the basis of regional tectonic arguments, which of these 200 Ma poles one chooses, esti­ Hamilton [1981] and Cordell [1982] suggest that mated paleolatitudes for North America are sig­ the Colorado Plateau experienced 3°-50 of nificantly different. All of the ''Late TriaBBic" clockwise rotation with respect to cratonic rocks May and Butlers Jurassic Apparent Polar Wander 11,525

TABLE 3. Reference Poles From the Colorado Jurassic-Early Cretaceous history of North Plateau After a 3.8° Clockwise American motion. Rotation is Removed In relation to paleomagnetic data, Francheteau and Sclater [1969] were perhaps the first to view Latitude Lon§~tude A95 APW paths in light of a punctuated equilibrium Pole °N deg. model for plate motion. Irving and Park [1972] similarly recognized that APW paths consist of Upper Morris 64.6 164.2 3.9 long, arcuate "tracks" separated by relatively Formation sharp "hairpins." Tracks were interpreted as Lower Morrison 58.6 146.2 4.2 periods of constant plate motion relative to the Formation magnetic pole, and hairpins were interpreted as Kayenta Formation 61.9 78.1 6.3 the record of periodic change in the direction of Wingate Formation 59.6 70.4 8.0 plate motion. Subsequently, the concept of hair­ Moenkopi Formation 55.4 106.5 5.3 pins received little attention in the late 1970s and early 1980s largely due to advent of the sliding-window technique for constructing APW paths that was popularized by Irving [1977]. east of the Rocky Mountains and the Rio Grande The recent analysis of Gordon et al. [1984] has Rift during Laramide and Neogene time. Gordon et revived the concept of hairpins, called "cusps," al. [1984] noted that paleopoles from rocks on and their recognition is potentially important to the Colorado Plateau are displaced systematically understanding the nature and implications of clockwise from equivalent age reference poles North American Jurassic APW. from other parts of North America. This is Paleopole data selected for the present analy­ especially evident with respect to the TriaBBic sis reveal two cusps within the Jurassic APW poles and can be interpreted either as a small path, an older cusp labeled "Jl" and a younger clockwise rotation of the plateau or as the cusp labeled "J2" (Figure 3). The apex of the Jl result of systematic errors in age assignments cusp is presently defined by the Wingate Forma­ and correlation. Steiner [1984] has suggested t ion pole and that of the J2 cusp by the lower that similar discrepencies are present between Morrison Formation pole (see the appendix). Each plateau and nonplateau poles of Pennsylvanian and of these cusps directly reflects a change in the Devonian age as well and argues for tectonic direction and velocity of North American plate rotation. Also noted by Steiner, however, is the motion, which in turn may be expressed on a lack of discordance between Permian poles. regional scale by episodes of intraplate deforma­ Bryan and Gordon [1985] quantitatively tion. Tracks separated by these cusps are analyzed the magnitude of potential Colorado labeled Tr-Jl, Jl-J2, and J2-K and represent Plateau rotation using paleomagnetic poles and intervals of North American plate motion found a clockwise value of 3.8°+2.9°. Using this describable by single poles of rotation. mean value, we have recalculated plateau poles The same tracks and cusps are recognizeable (Table 3, Figure 2b). The basic morphology of regardless of whether corrected or uncorrected the Triassic-Cretaceous APW path is unaffected by plateau poles are used. The Jl cusp is somewhat this recalculation, but differences in detail do modified because restoration of the Wingate and alter our interpretations of PEP analysis. Kayenta Formation poles decreases the "sharpness" of the cusp. Although the Sinemurian age Wingate pole still forms the apex of the Jl cusp, it is no longer statistically significant from poles As discussed originally by Francheteau and Sclater [1969] and more recently by Gordon et al. [ 1984], APW paths can be modeled as a series of small circle segments, each of which defines a paleomagnetic Euler pole (PEP) in the same way as do hot spot tracks or transform faults. The applicability of the PEP methodology is based upon the notion that large plates tend to rotate about one absolute motion pole for long periods of time (i.e., 107-108 years). Arguments in favor of plate motion stability include the long, continuous nature of fracture zones and the curvilinear nature of hot spot tracks [Gordon et al., 1984]. The generally accepted view of plate motion appeals to boundary conditions (i.e., ridges and trenches) as the primary control over both direc­ tion and velocity [Forsyth and Uyeda, 1975]. It follows that stable boundary conditions generate plate motions about single Euler poles at con­ stant angular velocity for long intervals of time. An alternative hypothesis would be that frequently changing boundary conditions should preclude plate motion of constant direction and Fig. 3. PEP model and terminology applied to the velocity. PEP analysis of the revised APW path revised North American Triassic-Early Cretaceous allows us to teat these two models for the APW path. 11 ,526 May and Butler1 Jurassic Apparent Polar Wander

a) NTII cordance as would a small clockwise rotation of 30 the plateau (see Summerville Formation pole in the appendix). UM 20 PEP Analysis NT! 10 The PEP technique employed to analyze the

0 distribution and geometry of the nine reliable Jurassic and Early Cretaceous reference poles was facilitated through use of a computer program -10 that allows small circles to be fit to a sequence of poles. The technique involves an iterative -20 minimization of the arc length from individual poles to the small circle plane used to fit the -30 trend of the poles. This program calculatea the w I/) ..I best fit PEP, the latitude of the small circle C( ::i about the PEP, a total residual (i.e., the aum of 120 100 eo ep 40 zp o• Q I I 1 I I individual pole versus small circle misfits), iii coordinates of individual data transformed into SMALL CIRCLE LONGITUDE Ill IC PEP space (i.e., rotated so that Euler pole is 10 coincident with north geographic pole), and

KA LM individual residuals associated with each pole. b) 0 UM G Residual values are simply the angular misfit of a pole with respect to the beat fit small circle. -10 Individual poles were not weighted according to their associated confidence parameters or to a -2 0 "standard error" as done by Gordon et al. [1984). cc 0 Most A95 are between 4° and s except for the two Newark Trend poles (1.4° and 2.3°). It was not Fig. 4. Residual distribution for (a) single PEP considered desirable to weight these latter two fit, and (b) double PEP fit. "Small circle poles heavily as there is some question whether longitude" is the coordinate of a pole trans­ or not these values are artificially small due to formed into PEP space and then standardized so overestimation of the actual number of indepen­ that the Wingate pole (W) is arbitrarily placed 0 dent sites (see the appendix). at o (uncorrected plateau pole data). In the present analysia, the Early Jurassic­ Early Cretaceous APW path is fitted with two whose ages range from Carnian-Norian to small circles rather than one, as was done by Plienabachian. The timing of the Jl cusp is Gordon et al. [1984). Because of their inter­ therefore not as distinct after correction for pretation of paleopole absolute age and because Plateau rotation and may be viewed as a 25-30 of inclusion of certain poles considered here to m.y. APW standstill interval of North American be unreliable (i.e., Summerville and Twin Creek plate motion reorganization. formations) (the appendix), they were unable to Because both the Wingate Formation and lower discriminate the two-track nature of the Morrison Formation poles are from rocks on the Jurassic-Cretaceous APW path. Our defenae of a plateau, the absolute arc length of the Jl-J2 two-track fit is based on simple visual inspec­ track is unaffected by rotation of the Colorado tion of the revised APW path as well as analysis Plateau. The J2 cusp as defined by the lower of the spatial distribution of residuals and a Morrison pole moves slightly to the east and to a trend line analysis ''F" test. lower latitude. The J2-K track becomes arcuate As shown in Figure 4, the distribution of with the opposite sense of concavity after rota­ residuals for a single PEP fit to the UPP data tion correction because two of the three poles set is not random along the track. The distribu­ defining this track are from the Morrison Forma­ tion is symmetrical with positive values at both tion on the plateau. ends (with the exception of W) and negative All of these modifications to the Jurassic APW values within the 116°-142°E longitude window path affect PEP analysis illustrating the sensi­ (60°-100° longitude relative transformed tivity of the latter technique for small data coordinates). The largest negative residual for sets. The following discussion addresses PEP this fit is associated with the lower Morrison parameters for two sets of Jurassic pole posi­ Formation pole, an observation we use to help tions1 one with uncorrected plateau poles {UPP) define the J2 APW cusp. This systematic distri­ (i.e., no tectonic correction) and one with bution implies failure of the single PEP model corrected plateau poles (CPP) (i.e., 3.8° to resolve structure inherent in the raw data clockwise rotation removed). Rotation of the set. Our single-track PEP at 85°N, 90°E is not Colorado Plateau is regarded with some suspicion significantly different from the "B" pole of because all of the poles from the plateau are Gordon et al. [1984) at 84°N, n°E. from sedimentary rocks which for the Triassic and The along-track distribution of residuals for Jurassic have a history of exhibiting present the single PEP fit of the CPP data set is some­ field secondary overprints which tend to bias what less obviously systematic, but again the pole locations toward the geographic north pole. largest negative residual is associated with the Because of the orientations of the Triassic and lower Morrison pole and positive residuals are Jurassic APW paths, small unremoved overprints of generated for both younger (Cretaceous average) this type provide the same general sense of dis- and older (Kayenta through Glance) poles (Table May and Butler1 Jurassic Apparent Polar Wander 11,527

TABLE 4a. Paleomagnetic Euler Pole Data1 PEPa for both UPP and CPP data sets are listed Uncorrected Plateau Poles in Table 4 and shown in Figures 5 and 6. The UPP Jl-J2 PEP falls within the North American plate in south-central Quebec while the CPP Jl-J2 PEP Pole Transformed Residual is located east of Florida. Both are shown with Latitude Longitude contoured solution spaces which represent the ON OE distribution of total residual as a function of location within the region of fit. In Flgurea 5 and 6 we show two contoured residual classes as J-K "Single Fit" labeled. PEP1 85.0°N, 90.0°E, small circle Solution spaces have elliptical shapes with latitude • 66. 7°, total residual • 13.8° long axes oriented perpendicular to the trend of the APW small circle, caused by the fact that the w 63.37 328.56 -3.346 azimuth of the APW track is better defined than K 66.75 336.33 0.033 the radius of curvature. The general shape of NTI 67.96 351. 77 1.243 the solution space for our beat fits is similar NTII 70.14 16.31 3.423 to that reported by Gordon et al. [ 1984). How­ cc 66.20 30.89 -.0513 ever, because we made no assumptions about the G 66.23 48.94 -0.484 predicted statistical description of the total LM 64.18 60.41 -2.536 residual distribution, we do not convert the UM 68.65 84.ll 1.932 contoured solution into a 95'%. conf iden~e field as KA 66.96 107.83 0.247 do Gordon et al. [1984). Although the absolute location of the beat fit PEPa for the Jl-J2 track is quite different Jl-J2 Track between the UPP and CPP cases, both poles lie PEP1 52.0°N, 286.0°E, small circle latitude • 26.5°, total residual • 5.1° TABLE 4b. Paleomagnetic Euler Pole Data1 w 26.33 156.93 -O. ll9 Corrected Plateau Poles K 27.57 162.01 l.ll5 NTI 26.39 168.67 -0.062 Pole 'rransformed Residual NTII 27.32 178.68 0.870 Latitude Longitude cc 24.08 185.16 -2.373 ON OE G 26.45 192.74 -0.003 LM 27.02 198.55 0.572 J-K Track "Single Fit" PEP1 85.0°N, 90.0°E, small circle latitude J2-K Track • 66.1°, total residual• 15.0° PEP1 31.0°N, 176.0°E, small circle latitude 52.5°. total residual - 0.09° w 64.26 336.99 -1.867 K 66. 77 345.74 0.644 LM 52.52 205.88 0.012 NTI 67.96 351.77 1.830 UM 52.46 188.76 -0.043 NTII 70.14 16.31 4.010 KA 52.54 173.86 0.031 cc 66.20 30.89 0.074 G 66.23 48.94 0.103 LM 61.ll 63.66 -5.016 4). The total residual for a single PEP fit to UM 65.52 84.79 -0.612 the CPP APW data is even larger than for the UPP KA 66.96 107.83 0.834 data (15.0° versus 13.8°). The location of the PEP is unchanged by use of the corrected plateau poles. Jl-J2 Track The single PEP residual distribution can be PEP: 30.0°N, 287.0°E, small circle latitude rectified by partitioning the data and using • 4.4°, total residual• 4.7° multiple trend lines. The Jurassic APW path is beat modeled as two tracks with an intervening w 4.56 162.38 0.141 cusp (J2) now recognized at approximately the 149 K 4.82 166.79 0.401 Ma lower Morrison pole. The distribution of NTI 4.92 169.40 0.506 residuals for the two PEP fit is nonayatematic, NTII 5.35 178.41 0.931 suggesting that this model better approximates cc 2.09 184.24 -2.326 the data distribution (Figure 4). We have also G 4.75 191.00 0.339 used a trend line analysis "F" teat to evaluate LM 4.42 199.29 0.008 the statistical significance in terms of total residual minimization afforded by the two PEP model over the one PEP model. This teat is J2-K Track significant at the 95'%. confidence level. The F PEP1 43.0°N, 22.0°E, small circle latitude teat is not significant at the 95'%. confidence • 21.6°, total residual• 0.01° level' if we increase the degrees of freedom further by adding a third segment to the Jl-K' APW LM 21.59 152.39 -0.027 path. It is not clear from visual inspection or UM 21.60 163.57 -0.0ll from residual distribution where a third track KA 21.65 173.62 0.037 would be fitted. 11,528 May and Butlers Jurassic Apparent Polar Wander

Fig. 5. Jl-J2 track poles with best fit PEPs1 (a) UPP data, (b) CPP data. Contoured solution space for total residuals• (a) 5.12°-6.13°, 6.13°-7.15° , (b) 4.65°-5.55°, 5.55°-6.46°. Shaded region of solution space in Figure 5b shows field of kinematical­ ly "reasonable" Euler poles. along the same great circle and in fact fall North American absolute motion must be located within each others optimum solution spaces. This north of Jurassic age Atlantic oceanic crust. reflects the sensitivity of PEP analysis to The shaded region of the PEP solution space in slight changes in curvature of an APW track and Figure 5b, or the UPP PEP (Figure 5a) are better emphasizes the difficulty in using PEPs to con­ estimates of a kinematically reasonable Jl-J2 strain location dependent parameters such as the track Euler pole. Again, this emphasizes the linear velocity for any point within the North poorly constrained distance of the PEP from the American plate. The best fit CPP PEP (Figure 5b) Jl-J2 track along the relatively well-constrained is unreasonable in that it does not describe the PEP great circle. North American rotation required to open the The effect of Colorado Plateau rotation is central Atlantic basin. Given the lack of sig­ even more dramatic for the J2-K PEPs. The UPP nificant APW for Africa during the Early-Middle J2-K PEP (Figure 6a) is located in the north Jurassic [Irving and Irving, 1982] and the hot central Pacific, whereas the CPP PEP (Figure 6b) spot model of Morgan [1983], any Euler pole for is located in Yugoslavia. Again, both PEPs fall

PEP ~

Fig. 6. J2-K track poles with best fit PEPs and small circle trends1 (a) UPP, (b) CPP. Contoured solution space for total residuals• (a) 0.09°-0.48°, 0.48°-0.88°, (b) 0.08°-o.34°, o.34°-0.61°. May and Butlers Jurassic Apparent Polar Wander 11 ,529

JURASSIC CRETACEOUS approximately constant boundary conditions will Early Mid die Late Early not only rotate about fixed Euler poles but should do so with constant angular velocities. 40 It is important to remember, however, that plate -+----*-----1 KA ---- velocities calculated in this way are minimum £ a) ,I estimates because of the longitude ambiguity I.LI ...J 20 inherent in paleomagnetic data. C) mean UM z angular velocity--.... Trend lines to the angular displacement data <( =34°/m y in Figure 7 are generated using unconstrained, 0 I.LI 40 0 unweighted linear regression. Correlation 0z coefficients of 0.98 and 0.99 for the UPP and CPP I.LI Jl-J2 regression lines, respectively, indicate bi mean angular velocity that these data are well described by a linear :::J 20 Cf) =07°/m.y fit and therefore by the PEP model. The passage of the UPP Jl-J2 velocity line through the lower Morrison data point in Figure 7a is solely a 0 coincidence of the linear regression fit (i.e., we did not constrain the line to necessarily 210 190 170 150 130 110 90 include this point). The J2-K track was con­ AGE (MA) strained to pass through both Morrison poles because (1) this is the true regression through a

data field of N•2 1 and (2) the fit is very poor JURASSIC CRETACEOUS if we include the oldest reliable Early Creta­ Earl Middle Lale Early ceous pole (Moteregian Hills 126±6 Ma) on this track (correlation coefficient • 0.90). Lacking other Tithonian-Berriasian age poles, the PEP b) model predicts a transition from J2-K rapid APW to the Cretaceous stillstand at approximately 140-142 Ma regardless of whether UPP or CPP poles £ -'--.~~------20 are used. The absolute age of this transition I.LI ...J 10 is, however, very dependent on the time scale we C) Z40 use to assign ages to the Morrison poles. If we <( 0 determine a best fit CPP Jl-J2 line excluding the 0 lower Morrison pole and then map the lower I.LI 0 20 mean angular Morrison pole subtended angle onto this line, an z velocity absolute age of 140 Ma is predicted. While not I.LI =06°/m' y I- Ill consistent with the Harland et al. [1982] time :::J scale, this estimate is consistent with the Van Cf) 0 Hinte [1976] time scale. Our application of the angular velocity 210 190 170 I 0 130 10 0 diagram differs from Gordon et al. [1984], who do AGE (MA) not use linear regression but constrain their lines to pass through data points corresponding Fig. 7. Angular velocity diagrams for (a) UPP to poles they interpret as track end points. But and (b) CPP Jurassic APW path. Subtended angles there is no reason why the position of the and best fit regression lines were calculated Wingate pole should be considered any better separately for the Jl-J2 and J2-K tracks and then determined than any other pole on the "J-K" composited. Values along right margin show sub­ track. Neither is there any reason why the tended angles for the J2-K track poles using the Monteregian Hills pole should represent the lower Morrison pole aa a reference. beginning of the Cretaceous stillstand. There­ fore the linear regreuion method is probably a approximately along a great circle perpendicular better approach to the angular velocity prob1em. to the J2-K track but rotation of the lower and The slopes of the Jl-J2 lines define angular upper Morrison poles causes the CPP J2-K track to velocities of 0.70/m.y. and 0.6°/m.y. for UPP and be concave northward as opposed to southward for CPP data sets, respectively. The J2-K lines, the UPP J2-K track. constrained only by the two Morrison Formation angular displacements, suggest dramatically Angular Velocity Analysis higher angular velocities of 3.4°/m.y. and 2.8°/m.y. during the Tithonian. We believe that The transformed coordinate data in Table 4 the increased plate velocity indicated by the can be used directly to evaluate the angular Morrison poles is real but that the absolute progression of poles along each of the two values suggested by PEP angular displacement Jurassic APW tracks. This is done by plotting analysis are approximate at best. the angular distance of successive poles (i.e., Angular velocities associated with each of the longitude along best fit circle) away from the PEPs yield linear velocities calculated at San Wingate Formation pole as a function of age. To Francisco, California, of about 5 cm/yr for the the extent that angular progression values UPP Jl-J2 pole and for a pole in the shaded approximate linear trends, we may conclude con­ region of the CPP Jl-J2 solution space (Figure stant angular plate velocity about a particular Sb). The UPP J2-K pole yields a linear velocity PEP. This is an expected corollary of PEP plate of about 30 cm/yr, and the CPP J2-K pole of 50 motion philosophy1 that is plates that experience cm/yr. These latter values can be somewhat 11 ,530 May and Butler1 Juraaaic Apparent Polar Wander reduced by selecting PEPa closer to San basins at 179±3 Ma [Sutter and Smith, 1979]. The Francisco, but atill within the calculated solu­ Jl cuap therefore correlates well with the ayn­ tion apace. However, even after selecting alter­ rift phase of Atlantic spreading history. Both native pole locations, linear velocities seem the age range of rift sediments and the apparent unreasonably high. Perhaps this reflects prob­ duration of plate motion reorganization aa recog­ lems with the age progreaaion of polea along the nized by the Jl cusp suggest a period of crustal J2-K track or inadequate constraint on the loca­ stretching of at least 25-30 11..y. tion of J2-K PEPa because of the small (Na3) data No significant APW ia recorded by 210-180 Ma aet. paleopolea from Africa and South America [Irving and Irving, 1982; Vilas, 1981], and the hot spot Implications for North American Plate Motions model of Morgan [1983] predicts relatively minor and Tectonics northwest motion of thea plates during the Early Jurassic. This requires that the opening of the One of the goala of APW analysis ia to under­ central Atlantic waa accommodated primarily by stand the kinematic history of plate motion in North American absolute plate motion [Steiner, relation to global tectonic and intraplate defor­ 1983]. This NW absolute motion of North America mational events. It is therefore instructive to during Jurassic opening of the central Atlantic investigate the potential correlation between is recorded by the Jl-J2 and J2-K APW tracks. abch events and the Jl and J2 cuapa of the The transition from rift to drift along the cen­ Jurassic APW path. The following diacuaaiona of tral Atlantic likely occurred during the late North American tectonics and Juraaaic APW are Early to Middle Jbraaaic (i.e., during the Jl-J2 baaed on the CPP path shown in Figure 2b. Thia track time). The linearity of the Jl-J2 track ia considered a more conservative and defensible segment on the angular velocity diagram (Figure approach because both Jl and J2 cuapa are defined 7) suggests that North American plate angular by plateau poles. Uncorrected, the Jl cusp is velocity was constant during this fundamental more pronounced, and one is enticed into more tectonic transition. elaborate tectonic scenarios than the true uncer­ tainties probably warrant.

Jl-J2 APW Track and Opening of the Atlantic Ocean Using the magnetic polarity time acale of Harland et al. [1982], the J2 cusp (late Correlations and relationships between the Kimmeridgian-early Tithonian) corresponds tem­ Jurassic APW path and the origin and evolution of porally with a change in orientation of central the central Atlantic Ocean have been discussed Atlantic 111arine magnetic anomalies at chron M21 for over a decade [Steiner, 1975; Dalrymple et time (Kimmeridgian) [Schouten and Klitgord, al., 1975; Smith and Noltimier, 1979]. We can 1982]. Associated with this plate reorganiza­ now compare the timing of the Jl and J2 cusps and tion, seafloor spreading may have ceased in the the direction of Jl-J2 and J2-K North American Gulf of Mexico leaving the Yucatan block and the motion with the Atlantic rift and drift history. Gulf basin as part of the northwestward moving The Sinemurian Wingate Formation pole is used North American plate. Continued spreading to define the Jl cusp, but at the 95'%. confidence between North America and South America occurred level it is not distinct from poles whose ages along a ridge system through the proto-Caribbean range from Carnian-Norian (Chinle Formation) to aouth of the Yucatan block [Pindell, 1985]. The Plienabachian (Kayenta Formation and Newark Group oldest marine magnetic anomalies recognized in I, 195+4 Ma) (Figure 5). Thia Late Triassic to the Venezuelan Basin as possible remnants of Early Jurassic timing of plate reorganization proto-Carribean spreading are approximately 150 corresponds temporally with the breakup of Pangea Ma [Ghosh et al., 1984]. and the separation of North America from Africa The major change in central Atlantic marine and South America. The syn-rift phase of this magnetic anomalies at chron M21 time, which we event is recorded by various Late Triassic-Early correlate with the J2 APW cusp, includes not only Jurassic red bed sedimentary sequences along the the general orientation but also geometrical North American Atlantic and Gulf coasts. detaiis. Between chron M21 and Mll time, Schouten The Chinle, Wingate, and Kayenta formations and Klitgord [1982] note the relative absence of are correlative with rocks in the Newark Super­ anomaly offsets, which implies to them that the group of the eastern United States [Olsen et al., Mid-Atlantic Ridge was relatively straight and 1982]. These sediments and interbedded lavas offset only by a few large transform faults. were deposited in fault-bounded basins inter­ Thia is consistent with rapid seafloor ap~eading, preted as pull-apart structures and half grabena which we suggest may correlate with rapid North formed during early rifting between North America American absolute plate motion during J2-K time. and Gondwana [Manapizer, 1981; Klitgord et al., Sundvik et al. [1984] suggest a primary seafloor 1984]. The oldest rocks in these basins are spreading origin for the transition from smooth considered to be Carnian in age [Olsen et al., to rough oceanic basement between anomaly Ml3 and 1982] indicating that some degree of baain Mll time. Thia is believed by them to reflect a development had begun by Late Triassic time. The decrease in the rate of mid-Atlantic spreading timing of actual plate separation and emplacement which may correspond to the transition from the of oceanic crust is not well constrained and has J2-K APW track to the Cretaceous atillatand. been estimated at anywhere from Plienabachian to Bathonian in age [Gradstein and Sheridan, 1983]. Intraplate Deformation in Western North America Klitgord et al. [ 1984] equate the initiation of seafloor spreading with the last major pulse of A number of tectonomagmatic events along the mafic igneous intrusions into the onshore rift western edge of North America correspond tempor- May and Butler1 Jurassic Apparent Polar Wander 11,531 ally with the Jl and J2 cusps of the Jurassic APW suggesting temporal correspondence with the Jl path. In the southern Cordillera the continental cusp. The second episode recognized by Oldow et magmatic arc active throughout the Early? and al. is the Late Jurassic Nevadan event, which Middle Jurassic in southeastern Arizona and produced a structural fabric coplanar with the northern Sonora shuts off at approximately 150 Ma older event and folding contemporaneous with west and sweeps quickly to the continental margin by to southwest vergent thrust or reverse faults. about 145 Ma [Coney and Reynolds, 1977; Damon et One implication of relating tectonic episodes al., 1981]. This may reflect a change in the observed in the western United States to cusps in angle of subduction of the downgoing Farallon the Jurassic APW path is that although the timing plate associated with a change in the relative suggests a cause/effect relationship, the change and/or absolute plate motion across the southern in North American absolute plate motion is not Cordilleran trench. This implied reorganization reflected by different orientations of structures corresponds closely with our estimated age of the in the Sierran region. As discussed by Beck J2 cusp and the associated change in North [1983] and Moore and Karig [1980], the axis of American absolute motion. A similar westerly arc shortening in zones of oblique subduction repre­ migration is recognized at about this time in sented by structures in the leading edge of an northern California [Saleeby et al., 1982]. upper plate la not necessarily parallel to the As had been previously discussed by Steiner vector of relative plate motion. Rather, the [1978, 1983] and Kluth et al. [1982], Gordon et shortening axis probably reflects the normal al. [1984] recognized a subinterval of rapid APW component of convergence and la constrained by in "Middle and Late Jurassic time" documented by the shape and orientation of the plate boundary. the Summerville and Morrison Formation poles. In other words, although episodes of intraplate The latter authors note a possible correlation deformation may be the result of changes in plate with the Nevadan Orogeny in the northern Sierra motion and therefore correspond to APW cusps, the Nevada and Klamath Mountains, suggesting that expression of accommodated strain as reflected this event may be more closely related to North within regional structural fabric may bear no American absolute motion than to relative plate relation to the direction of motion of either motions. We agree in part with this interpreta­ plate. The temporal correlations of the above tion. However, recognition of the J2 cusp makes outlined tectonomagmatic events in western North the temporal coincidence significantly more America with the Jl and J2 cusps are intriguing appealing and the plate kinematic scenario more and should be noted in regional tectonic consistent with PEP philosophy. syntheses. The Nevadan Orogeny has been interpreted by Saleeby et al. [1982] as an event of crustal Terrane Displacement shortening associated with collapse of an interarc basin within which oceanic crust as The western Cordillera of North America is a young as 157 Ma was being generated. In their collage of tectonostratigraphic terranea whose model, approximately E-W directed convergence paleogeographic relationships during Paleozoic caused thrusting of older Mesozoic arc rocks on and Mesozoic time with respect to cratonic North the east over interarc basin rocks on the west, America and to each other are uncertain [Coney et with associated incorporation of ophiolites and al., 1980; Coney, 1981]. Paleomagnetic data intrusion of peridotitic to dioritic igneous from these terranes have proven useful for complexes. The timing of the Nevada Orogeny in quantifying latitudinal displacements and the northern Sierra Nevada and Klamath Mountains azimuthal rotations [Beck, 1976, 1980]. Tectonic is well constrained at about 145-150 Ma. The translation/rotation is measured by comparing youngest strata affected by Nevadan deformation observed paleomagnetic directions with expected are Kimmeridgian and a Pb/U zircon date from the directions calculated from cratonic reference predeformation or syndeformation Bear Mountain poles. For this reason, reliable cratonic pluton in the Klamath& la 149+2 Ma [Saleeby et reference poles are fundamental to accurate al., 1982]. Published dates for the lower Coon estimation of the displacement history of suspect Mountain intrusive complex which postdates the terranes within orogenic belts. The Late Nevadan cleavage range from 142 to 150 Ma. The Triaasic through Jurassic APW path presented in apparent age of the Nevadan Orogeny therefore this paper differs significantly from previous correlates very well with the J2 cusp and atten­ APW analyses and from various reference poles dant change in North American absolute plate previously calculated for comparison with motion. Paleomagnetic data interpreted as a specific paleomagnetic studies of Cordilleran ''Nevadan" remagnetization have been obtained from terranes. rocks in the northern Sierra Nevada [Bogen et Paleomagnetic data from Late Triassic and al., 1985]. A paleomagnetic pole calculated from Jurassic rocks in a number of suspect terranes these results is indistinguishable at the 951 are discussed in light of the revised APW path. confidence level from either of the Morrison Cratonic reference poles appropriate for each Formation poles. study have been calculated with Fisher statistics Oldow et al. [1984] recognize three major assigning unit weight to each paleopole (Table episodes of structural deformation in Mesozoic 5). In cases where only two pole• are averaged, arc rocks of the northern Sierra Nevada and the larger of the two confidence parameters is northeastern Oregon. The oldest of these la a taken for the average pole. This may be a some­ Triassic-Jurassic event expressed as iaoclinal what conservative approach, but recalculation of folds with steep N-NW striking axial plane reference poles from individual weighted VGPa cleavage, steep fold axes and steep lineations. does not alter the conclusions. In many cases, Radiometric dates associated with these struc­ recalculation of concordance/discordance values tures are approximately 200 Ma [Saleeby, 1981], merely modifies the magnitude of apparent dis- 11'532 May and Butler: Jurassic Apparent Polar Wander

TABLE 5. Cratonic Reference Poles for Concordance/Discordance Calculations

Age, Latitude Longitude A95 No. Pole Age Ma ON OE deg.

1 Manicouagan + Chinle Late Triassic 220 58.4 84.4 7.0

2 Chin le Formation Carnian-Norian 220-230 57.7 79.l 7.0

3 Newark Group I Ar/Ar 195 63.0 83.2 2.3

4 Wingate + Kayenta + Sinemurian- Newark Group I Pliensbachian 200 61.6 77.0 5.4

5 Wingate Formation Sinemurian 200-206 59.6 70.4 8.0

6 Ka yenta Formation Pliensbachian 194-200 61.9 78.1 6.3

7 Wingate + Kayenta Sinemurian- Pliensbachian 200 60.8 74.1 8.0

See footnote for Table 1. (Note that Wingate and Kayenta Formation poles used in this table are corrected for plateau rotation.) placement, in other cases previously discordant implies that there are no significant inclination results become concordant, and, in at least one anomalies for Stikinia and Quesnellia during Late case, a previously concordant result becomes Triassic and Early Jurassic time. discordant. Characteristic components of the Stikine Directional concordance/discordance was calcu­ terrane include Mississippian and Permian lated according to the technique of Beck [ 1980] volcaniclastic, volcanic, and carbonate sedi­ as modified by Demarest [1983] or by converting mentary rocks [Coney, 1981]. This upper Paleo­ observed and expected inclination data into zoic submarine volcanic arc assemblage is over­ paleolatitudinal bands. Important in such calcu­ lain by Late Triassic through Middle Jurassic lations and their attendant interpretation is the volcanic rocks, predominantly basaltic and realization that even the best constrained values andesitic subaerial flows and pillow lavas generally provide a latitudinal resolution of no [Monger et al., 1982]. Paleomagnetic data from better than 500-800 km. Stikinia come from the Takla and Hazelton Group Cratonic reference poles which include or are volcanics of this latter ~ssemblage [Monger and based on Colorado Plateau poles were calculated Irving, 1980]. with both CPP and UPP data sets. Since these The Takla Group volcanics of Stikinia are data sets are very similar, it makes no differ­ Carnian-Norian in age [Monger and Irving, 1980] ence to conclusions of concordance or discordance or approximately 220-230 Ma. The most appro­ which is used, and thus only corrected plateau priate reference pole for this study is the pole values are shown in Table 6. Chinle Formation pole of Reeve and Helsley [1972] given in Table 5. The age of the Chinle Formation Terrane 1 also is Carnian-Norian, based on correlations to Newark Group strata by Olsen et al. [1982]. Stikinia and Quesnellia along with the Cache Alternatively, we can compare the Late Triassic Creek terrane and the Eastern Assemblage comprise Stikine results with a reference pole based on an the inboardmost "superterrane" of the Canadian average of the 215±5 Ma Manicouagan pole and the Cordillera which is tectonically juxtaposed on Chinle Formation pole. Thia average pole the east with Paleozoic miogeoclinal strata (58.3°N, 84.4°E) and the Chinle pole are both (Figure 8). Recent paleomagnetic studies from very different from the Late Triassic reference rocks of Stikinia and Quesnellia have been inter­ pole used by Monger and Irving [1980] (68°N, preted as evidence for significant latitudinal 93°E). displacement of these terranes with respect to Monger and Irving [1980] and Irving et al. cratonic North America [Monger and Irving, 1980; [ 1980] presented paleomagnetic data from Tak la Symons, 19831 Symons and Litalien, 1984]. Group volcanic rocks exposed along the eastern Apparent discordance between observed and side of Stikinia. Samples were collected from 14 expected inclinations from Late Triassic and sites at two localities (Asitka Peak and Sustut Early Jurassic rocks seemed to indicate approxi­ Peak, Table 6) within the Savage Mountain Forma­ mately 1500 km of northward relative motion t ion, a thick sequence of predominantly pillow during the late Mesozoic and early Cenozoic. lavas and subaerial basalt flows of late Carnian Reconsideration of these paleomagnetic data in to earliest Norian age (Figure 8). The mean light of the Harland et al. [1982] time scale and direction from the Asitka Peak locality differs our revised list of reliable cratonic reference by 19° in declination and 6° in inclination from poles leads to an alternative scenario regarding the mean at the Sustut Peak locality (Table 6). terrane displacement. Most importantly, the The difference in declination may reflect a small revised set of cratonic reference poles (Table 5) amount of relative rotation. The difference in TABLE 6. Concordance/Discordance Data: Terrane 1

AS, +s, Do, Io, ~ 5 , Reference DX, Ix, ~5' R±t.R, F±t.F, Rock Unit Age Ma ON OE deg. deg. deg. Pole deg. deg. deg. deg. deg. Reference

Takla Group Asitka Peak Carnian-Norian 225 56.7 234.6 300 44 6 1 342.9 46.1 7.0 -42.9± 9.0 2.l:t 8.1 A Asitka Peak Carnian-Norian 225 56.7 234.6 300 44 6 2 345.7 44.4 7.0 -45.7± 9.0 0.4:t 8.4 A Takla Group Sustut Peak Carnian-Norian 225 56.6 234.5 281 38 7 1 342.8 46.0 7.0 -bl.B:t 9.3 8.0:t 8.7 A ! Sustut Peak Carnian-Norian 234.5 7 345.7 44.3 7.0 -64.7± 9.2 ..1:1 225 56.6 281 38 2 6.3± 8.8 A c:i. Takla Group "'c: Averase Carnian-Norian 225 41 6 1 46 7 5 :t -8 A ..." Average Carnian-Norian 225 41 6 2 43 7 2 :t -8 A • Hazelton Group "' Nilkitkwa Formation Toarcian -190 55.6 234.6 359 55 16 3 345.4 49.9 2.3 13. 6±22. 5 -5.1±12.6 A ~ Hazelton Group .."' Ill Telkwa Formation 1 late Sinemurian 200-203 55.8 234.4 242 56 18 4 348.0 47.6 5.4 -106.0±25.9 -8.4±14.5 A ....Ill Telkwa Formation 1 late Sinemurian 200-203 55.8 234.4 242 56 18 5 351.l 44.4 8.0 -109.1±26.4 -11.6±15.7 A n Telkwa Formation 1 late Sinemurian 200-203 55.8 234.4 242 56 18 7 349.3 46.3 8.0 -107.3±26.3 -9.7±15.6 A .,,> Hazelton Group .. Telkwa Formation 2 late Sinemurian 200-203 56.5 234.2 294 52 25 4 347.8 48.5 5.4 -53.8±33.3 -3.5±19.6 A ::1:1 Telkwa Formation 2 late Sinemurian 200-203 56.5 234.2 294 52 25 5 350.9 45.3 8.0 -56.9±33.7 -6.7±20.4 A Telkwa Formation 2 late Sinemurian 200-203 56.5 234.2 294 52 25 7 349.1 47.2 8.0 -55. l:t33. 7 -4.8±17.3 A ""Cl ...0 Guichon Batholith K/Ar -200 50.5 239.0 28.3 36.3 7.3 4 350.8 40.4 5.4 37.5± 8.5 4.l:t 8.2 B .. Guichon Batholith K/Ar -200 50.5 239.0 28.3 36.3 7.3 5 353.9 36.8 8.0 34. 7± 9.9 0.5±10.9 B "' ~ Guichon Batholith K/Ar -200 50.5 239.0 28.3 36.3 7.3 6 350.3 40.9 6.3 38.0± 9.0 4.b± 8.9 B 1:1 Guichon Batholith K/Ar -200 50.5 239.0 28.3 36.3 7.3 7 352.1 38.9 8.0 36.2± 9.9 2. b:tlO. 6 B •c:i. Copper Mountain Intrusions K/Ar -200 49.3 239.4 25.9 41.2 3.6 4 351.l 38.6 5.4 34 .8:t 5. 9 -2.6± 6.7 c "' Copper Mountain Intrusions K/Ar -200 49.3 239.4 25.9 41.2 3.6 5 354.l 35.0 8.0 31.8± 7.7 -6.2±10.U c Copper Mountain Intrusions K/Ar -200 49.3 239.4 25.9 41.2 3.6 6 350.6 39.2 6.3 35.3± 6.6 -2.0± 7.5 c Copper Mountain Intrusions K/Ar -200 49.3 239.4 25.9 41.2 3.6 7 352.4 37.l 8.0 33.5• 7.7 -4.1• 9.6 c

AS, site latitude; +s, site longitude; D , declination observed; 1 , inclination observed; ~ , Fisher statistic; Reference pole, Table 5; 0 0 5 Dx, declination expected; Ix, inclination expected; R, rotation; F, flattening; References: A, Monger and Irving [1980]; B, Symons [1983]; C, Symons and Litalien [1984].

IJ1 VJ VJ 11'534 May and Butlers Jurassic Apparent Polar Wander

present latitudinal position with respect to cratonic North America during the Late Triassic. The mean declinations from the Takla Group volcanics are clearly discordant with "R" values of -43° to -65°. This rotation probably reflects local block rotation along the eastern side of Stikinia rather than wholesale rotation of the entire Terrane 1 composite. The Hazelton Group volcanics of Stikinia are divided into two age groups of late Sinemurian and mid-Toarcian ages by Monger and Irving [1980]. The former with an approximate age of 200-203 Ma may be compared to reference poles calculated from the Wingate pole, the Wingate + Kayenta pole, and the Wingate + Kayenta + Newark Trend Group I pole. The mid-Toarcian group has an approximate age of 190 Ma, and thus may be compared to the Newark Group I pole. Each of these reference poles is different from the reference poles used for the Hazelton results by Monger and Irving [ 1980] (i.e., 60°-63°N versus 78°N present latitude). Monger and Irving [1980] and Irving et al. [1980] present paleomagnetic data from the late Sinemurian Telkwa Formation and the mid-Toarcian Nilkitkwa ForDLation of the Hazelton Group in eastern central Stikinia (Figure 8). Four sites at each of two localities were collected from subaerial basalt flows and fine grained tuffs of the Telkwa Formation. Expected directions and concordance/discordance calculations based on our reference poles (Table 6) illustrate that irre­ spective of the reference pole selected, both localities have concordant inclinations. Thirty­ seven cores from seven sites were studied by Fig. 8. Suspect terrane map of northwestern Monger and Irving [1980] from the Nikitkwa Forma­ North America, modified from May et al. [1983]. tion in the Bait Range (Figure 8). Comparison Terranes include Ac, Alexander (Craig of the observed mean direction with an expected subterrane); Aa, Alexander (Admiralty direction calculated from the Newark Trend Group subterrane); An, Alexander (Annette subterrane); I pole reveals concordance in both declination c, Chugach; W, Wrangellia; S, Stikinia; Cc, and inclination (Table 6). At the 95% confidence Cache Creek; QN, Quesnellia; YT, Yukon-Tanana; level, Terrane 1 was situated at its present and EA, Eastern assemblages. Paleomagnetic latitude relative to North America during the localities1 1, Takla Group, Asitka Peak; 2, Takla Early Jurassic based on paleomagnetic data from Group, Sustut Peak; 3, Hazelton Group, Nilkitkwa the Hazel ton Group. However, owing to the very Formation; 4, Hazelton Group, Telkwa Formation 11 large 85 values for the three studies (16°, 18°, 5, Hazelton Group, Telkwa Formation 2; 6, and 25 ), the latitudinal resolution of the Guichon Batholith; 7, Copper Mountain Intrusions; Hazelton data is quite limited. 8, Hound Island volcanics; 9, Alaska Range; 10, Quesnellia is characterized by upper Paleozoic Wrangell Mountains; 11, Karmutsen volcanics, and Lower Triassic volcanic, volcaniclastic, and Vancouver Island; 12, "Seven Devils", southeast carbonate rocks and like Stikinia is interpreted Oregon. TF, Tintina fault; RMT, Rocky Mountain as a submarine volcanic arc assemblage [Coney, Trench. 1981]. Quesnellia also contains Upper Triassic and Lower Jurassic volcanic and elastic rocks which are intruded by Late Triassic-Early inclination is largely due to an obvious outlier Jurassic quartz dioritic to granitic plutons. in the Sustut Peak data set [Monger and Irving, Paleomagnetic data from Quesnellia have been 1980, Figure 4]. One of the eight sites from reported from two of the early Mesozoic plutonic this locality has a WSW declination and an complexes (Guichon Batholith and Copper Mountain anomalously shallow inclination, which produces a Intrusions) [Symons, 1983; Symons and Litalien, low k value (18) and biases the locality mean 1984]. direction toward a shallower inclination. Exclu­ The Guichon Batholith and the Copper Mountain sion of this single outlier from the Sustut Peak Intrusions of Quesnellia are both dated at locality brings the two locality mean directions approximately 198 Ma [Symons, 1983; Symons and into much closer agreement. Regardless, the mean Litalien, 1984]. Four alternative reference inclinations at both localities as well as the poles are used for comparison with these two formation average inclination of 41° given by Terrane 1 poles in Table 61 (1) Wingate Formation Monger and Irving [1980] are concordant with pole, (2) Kayenta Formation pole, (3) Wingate + expected directions predicted by either reference Kayenta pole, and (4) Wingate + Kayenta + Newark pole listed in Table 5. The simplest interpreta­ Trend Group I pole, all with estimated ages in tion is that within the resolution provided by the range of 194-206 Ma. Listed in Table 5, these paleomagnetic data, Terrane 1 was situated at its four alternative reference poles are not May and Butler1 Jurassic Apparent Polar Wander 11'535 statistically distinct, and it makes very little American affinities, one possible scenario that difference to concordance/discordance calcula­ will accommodate these various paleomagnetic tions which reference pole one chooses. Each of data places Terrane 1 at roughly its present them ia, however, significantly different from latitude with respect to the craton in the Late the reference pole (70°N, 94°E) chosen by Symons Triassic-Early Jurassic followed by southward [1983] and Symons and Litalien [1984] for com­ translation until the ?Late Jurassic. Plate parison with the Guichon and Copper Mountain motion models of Engebretson [1982] show a change results. at about 145-150 Ma from SE to NE convergence of A mean direction of magnetization baaed on 49 the 'Kula plate with respect to North America. specimens from 13 aitea in the Guichon Batholith Unfortunately, there are no reliable Middle or [Symons, 1983] (10 normal, 3 reversed) ia concor­ Late Jurassic paleomagnetic poles from Terrane 1 dant in inclination but discordant in declination to test this south then north displacement model. with respect to all reference directions shown in Craton-derived detritus in the Bowser Basin indi­ Table 6. Approximately 35°-38° of clockwise cates that Terrane 1 was juxtaposed with the rotation may be related to distributed shear craton by the Early Cretaceous [Eisbacher, 1974; "ball-bearing" style tectonics associated with Monger et al., 1982]. Alternatively, one might dextral atike-alip along the southern Rocky question the tectonic applicability of these Mountain Trench system. The observed inclination paleomagnetic data from rocks with little or no for the Guichon Batholith differs by only 0.5°- paleohorizontal control. It ia very likely that 4.60 from expected 200 Ma inclinations. The rocks in the Coast Plutonic complex of British same basic results are obtained if we compare the Columbia have experienced post-Middle Cretaceous Guichon data with older Manicouagan or tilting in this region of complex Late Cretaceous Manicouagan + Chinle reference poles. and Tertiary deformation. Furthermore, the trend Symons and Litalien [1984] report new paleo­ of a horizontal rotation axis that will moat magnetic data from the Late Triassic-Early simply explain the discordant paleopoles from Jurassic Copper Mountain Intrusions of southern rocks such as the Spuzzum Pluton [Irving at al., British Columbia. Like the Guichon Batholith, 1985] is approximately parallel to the dominant the Copper Mountain granitoida were emplaced into regional structural trend (i.e., NW-SE) and would the Late Triassic volcanic and sedimentary Nicola require modest tilt to the southwest. Group rocks of Quesnellia (Figure 8). Symons and Litalien [1984] suggest an age of 198 Ma. Irre­ Terrane 2 spective of the specific Early Jurassic reference pole chosen, concordance/discordance results are Wrangellia ia a well-known tectonostrati­ similar to the Guichon values. Positive "R" graphic terrane, fragments of which are now values indicate approximately 320.35o of recognized from NE Oregon to SE Alaska. It is clockwise rotation, while "F" values are all characterized by a Pennsylvanian and Early concordant (Table 6). Once again, one must con­ Permian andesitic arc sequence overlain by Middle clude on the basis of concordant inclination from and/or Late Triassic tholeiitic basalt flows and the Copper Mountain Intrusions that Queanellia pillow lavas, in turn overlain by Late Triassic was located at its present latitude relative with platform carbonates [Jones et al., 1977]. A to cratonic North America during the Early number of paleomagnetic studies of Triassic vol­ Jurassic. canic rocks from Wrangellia have been interpreted In summary, all of the available paleomag­ to show latitudinal displacements of up to 3000 netic data from Late Triassic and Early Jurassic km with respect to the craton. The magnitude of rocks of Terrane 1 have concordant inclinations paleomagnetic discordance from Wrangellia is when compared to appropriate, reliable reference modified by the revised Late Triauic-Jurauic directions from cratonic North America. This APW path. conclusion ia in direct conflict with the inter­ Paleomagnetic investigation of Middle-Late pretations of Monger and Irving [1980], Symons Triassic basalts from four different regions of [1983], and Symons and Litalien [1984]. These Wrangellia all predict approximately the same latter workers suggested as much aa 100-15° of paleolatitude even though these fragments are now relative latitudinal displacement between Terrane distributed over 2500 km along the Cordilleran 1 and North America in late Mesozoic and early margin. The moat appropriate reference pole with Cenozoic time. which to compare Late Triassic Wrangellia Concordant Late Triassic and Jurassic paleo­ results is the Carnian-Norian age Chinle Forma­ magnetic results from Stikinia and Queanellia tion pole (Table 5). Reference poles for seem to be in conflict with geological and paleo­ Wrangellia studies have consistently been chosen magnetic data from younger rocks in the same at significantly higher latitudes and more terranes. These data seem to suggest 1000-2500 easterly longitude11 64°N, 92°E [Hillhouse, km of post-Middle Cretaceous northward transla­ 1977], 68°N 1 93°E [Yo le and Irving, 1980], tion outboard of the Tintina•Rocky Mountain 65.3°N, 94.2°E [Hillhouse et al., 1982] 1 and Trench fault system. Paleomagnetic data of Rees 61.4°N, 92.5°E [Hillhouse and Gromme, 1984]. et al. [1985] and Irving et al. [1985] have been These reference poles suffer from including both interpreted to indicate as much as 2000-2500 km unreliable data and numerous results from rocks of post-Middle Cretaceous northward translation now known to be of Early Jurassic age. For of Terrane 1. Thia conclusion ia based on pri­ example the reference pole of Hillhouse and mary magnetizations observed in Cretaceous batho­ Gromme [1984] is calculated from nine paleopolea, li ths of the Coast Plutonic Complex and on secon­ only two of which are from Late Triassic rocks. dary magnetizations observed in Late Triassic and The reliability of the pole from the Watchung Jurassic rocks of Stikinia. Basalts [Opdyke, 1961] baa recently been Because late Paleozoic and early Mesozoic questioned by Mcintosh et al. [1985] (see the fossils from rocks in Terrane 1 indicate North appendix) and is Early Jurassic in age. Similar- 11 ,536 May and Butler: Jurassic Apparent Polar Wander

TABLE 7. Paleolatitude Results From Wrangellia

Locality Observed Expected Poleward Displacement, Paleolatitude Paleolatitude km ON ON

Wrangell Mountains, Alaska [Hillhouse, 1977] 7.7-13.1 27.3-40.8 2541*

Alaska Range, Alaska [Hillhouse and Gromme, 1984] 10.1-17.7 29.1-43.4 2387*

Vancouver Island, British Columbia [Yole and Irving, 1980] 15.3-19.3 15.1-24.5 253*

Southeast Oregon, "Seven Devils," [Hillhouse et al., 1982] 13.4-21.6 9.7-17.9

*Northern hemisphere interpretation. tNorthern hemisphere, south translation. ly, poles from the Newark Trend intrusive and One of the more enigmatic paleomagnetic extrusive rocks are known to be Early Jurassic, results from the western Cordillera has been the as is the Kayenta Formation pole. On the other Hound Island Volcanics pole published by hand, the Moenkopi Formation, from which two Hillhouse and Gromme [1980]. The Hound Island poles were used by Hillhouse and Gromme, is Early Volcanics are part of a submarine volcanic arc Triassic and significantly older than the assemblage consisting of pillow basalts, pillow Wrangellia basalts. The reference pole of breccias, andesitic breccia, aquagene tuffs, and Hillhouse and Gromme therefore includes poles minor limestones assigned to the Admiralty sub­ whose ages range from 245 to 195 Ma and which terrane [Hillhouse and Gromme, 1980]. As origi­ fall on both the Tr-Jl and Jl-J2 APW tracks of nally interpreted, the Hound Island paleomagnetic the North American APW path. pole was concordant with respect to a North We have compared the observed paleolatitudes American Late Triassic reference. However, this for Wrangellia compiled by Hillhouse and Gromme apparent concordancy is simply an artifact of the [1984] with expected paleolatitudes based on the reference pole used. Hillhouse and Gromme [1980] Chinle reference pole (Table 7). Inclinations use a "Late Triassic" cratonic pole at 65.3°N, from both the northeastern Oregon (Seven Devils 94.2°E, which is based entirely on paleopoles and Huntington Arc rocks) and the Vancouver from Early Jurassic Newark Trend igneous rocks Island (Karmutsen Volcanics) (Figure 8) are con­ and from the Early Jurassic Kayenta Formation. A cordant, suggesting no significant latitudinal more appropriate reference pole is either the displacement of these fragments since the Late Chinle Formation pole of Carnian-Norian age or an Triassic. Inclinations from the Wrangell average Chinle + Manicouagan pole with a somewhat Mountains and Alaska Range (Nikolai Greenstone) younger average age (Table 5). The Hound Island (Figure 8) are discordant and suggeit approxi­ observed direction is discordant in both declina­ mately 2300-2500 km of northward latitudinal tion and inclination with respect to expected displacement, consistent with their present geo­ values calculated from either reference pole. graphic separation from the southerly fragments Rotation values suggest counterclockwise rotation of Wrangellia. Assuming a northern hemisphere of 110°±30°, while negative flattening values can location in the Late Triassic, concordance/dis­ be interpreted either as evidence for southward cordance calculations based on the revised translation of the Alexander terrane or extreme reference pole suggest that all of Wrangellia northward translation from the southern was located at the approximate latitude of hemisphere. In the Late 'rriassic the Alexander northern Oregon-southern British Columbia. terrane was located at approximately 47° either Expected paleolatitudes were also calculated north or south, while-the expected North using a reference pole (57.3°N,90.1°E) which is American latitude was 21°-29° north. the average of the Chinle and Moenkopi poles. This reference pole is appropriate for rocks Conclusions slightly older than the Carnian-Norian Chinle Formation and allows for the fact that the The Jurassic APW path for North America pre­ Nikolai Greenstone may be as young as Ladinian. sented in this paper differs from previously The resulting paleolatitudes are all Within 1.6° published paths generated with various smoothing of those listed in Table 7 so the basic conclu­ techniques. The new path generally predicts more sions are unaffected by choice of either Chinle southerly paleolatitudes for North America than or Chinle+Moenkopi reference poles. do any of several APW paths now in use. Belief May and Butlers Jurassic Apparent Polar Wander 11 ,537 in the accuracy of our selected data base allows on the basis of an unacceptable k value.), (4) confident recognition of APW cusps in the Late a95 ~ 15°, (5) age known to within :!:10 m.y. Triassic-Early Jurassic and Late Jurassic (Jl and (assuming the absolute precision of the Harland J2). Cusps and intervening tracks are well et al. [1982] time scale), and (6) sufficient described by PEP modeling and indicate periods of discussion of geologic setting such as to demon­ constant velocity of North American plate motion strate an appropriate understanding of necessary from the Sinemurian to early Tithonian and from structural corrections. the early Tithonian to the early? Berriasian. The timing of the Jl and J2 cusps corresponds Late Jurassic: 144-163 Ma with Atlantic Ocean and western North America tectonic events. Finally, PEP analysis Three paleopoles from Late Jurassic rocks of accentuates the spatiotemporal relationships of North America are considered reliable. Two poles reliable cratonic paleopoles, allowing more con­ from the Morrison Formation in Colorado and one fident selection of reference poles for terrane from the Glance Conglomerate in southeastern displacement studies. Revised Late Triassic­ Arizona record APW during late Kimmeridgian and Early Jurassic reference poles for North America Tithonian time (Figure 2). No reliable poles are indicate different amounts of tectonic transport available from Oxfordian or early Kimmeridgian for some western Cordilleran terranes than have rocks. been previously proposed. Previously published Morrison Formation. The youngest reliable paleomagnetic data from Stikinia and Quesnellia Jurassic poles for cratonic North America are yield concordant inclinations implying that these from the upper and lower Morrison Formation of terranes were at approximately their present the Colorado Plateau (Table 1). Steiner and relative latitude with respect to cratonic North Helsley [1975] studied the polarity stratigraphy America during the Late Triassic-Early Jurassic. of a 165-m-thick section of Morrison sandstone Southern fragments of Wrangellia also yield con­ and mudstone near Norwood, Colorado, and Steiner cordant inclinations with respect to revised Late [1980] discussed further results from this same Triassic reference poles if a northern hemisphere locality. The 425 samples collected at an origin is assumed. average spacing of 30 cm were subjected to pro­ gressive thermal demagnetization to 66o 0 c. Appendix Approximately 50% of these samples retained a large component of Brunhes field overprint even In this appendix, paleomagnetic poles from after thermal cleaning so that a selected data Jurassic rocks of North America are reviewed. set of only well-behaved sample directions was Discussion of reliable poles is given first, used to calculate pole positions. followed by brief explanations of why certain Steiner and Helsley [1975] calculate poles in poles that have been used in past compilations a variety of different ways. Poles for were rejected. individual polarity zones segregate into two Reliability criteria employed in the present stratigraphically distributed clusters: a lower analysis include (1) demonstration that a stable, group Rl, R4, and N2 and an upper group RS, R6, primary component of magnetization was isolated R7, and NS. We have converted polarity interval through standard alternating field (af) and/or mean directions to VGPs for averaging, recog­ thermal demagnetization techniques (preferrably nizing these are not true VGPs but probably are both), (2) N > 10 (number of sites = number of viable paleomagnetic poles themselves. Thus the independent geomagnetic field readings, each formation mean poles (upper and lower) can be preferrably determined from multiple samples), viewed as averages of multiple Morrison poles. (3) Fisher precision parameter associated with Both subsets of poles pass the reversals test at VGP dispersion, 20< k <150 (Important criteria 95% confidence. These considerations do not for acceptance of a paleomagnetic pole include affect the mean pole positions but do affect the assurances that secular variation of the geomag­ associated confidence regions. That is, the netic field has been averaged and that secondary radius of the confidence circle is probably components of magnetization have been removed. overestimated. These criteria are in part evaluated on the basis Following Steiner and Helsley [1975], the of observed "k" values that reflect the disper­ difference in pole positions for the upper and sion of directions or VGPs. Models concerning lower Morrison Formation is interpreted to paleosecular variation including expected disper­ reflect APW during Morrison deposition. The sion values have been discussed by Cox [1970], outcrop expression of this apparent temporal McElhinny and Merrill [1975], and Merrill and separation is unclear, but it may reflect a McElhinny [ 1983]. Very high k values (i.e., hiatus between the Salt Wash and Brushy Basin k)l50) often indicate failure to average secular members of the Morrison Formation at the Norwood variation (except in slowly cooled plutonic locality. terrains), whereas very low k values (i.e., k<20) The biggest problem associated with the suggest the presence of uncleaned secondary com­ Morrison poles is the controversial age of this ponents causing greater dispersion than expected formation. In their original paper, Steiner and from paleosecular variation. Another cause of Helsley [1975] cite the ''Portlandian" affinity of low k values is improp~r tilt correction which the dinosaur fauna and the pre-Purbeckian, post­ can be important in studies of mildly deformed Oxfordian estimation of Imlay [1952]. Steiner volcanic sequences [Beck and Burr, 1979]. The [1980] again cites Imlay [1952] but this time for limiting values chosen for this analysis are a mid-late Oxfordian age for the lower part of somewhat arbitrary and are only used in con­ the Morrison near Norwood, Colorado. Although junction with other data such as a 95 and demag­ not explicitly stated, the main reason for netization results. No study was rejected solely suggesting this older age for the Morrison seems 11 , 538 May and Butler1 Jurassic Apparent Polar Wander to be Steiner's [ 1980] belief that the observed tionally on the "middle and upper" members of polarity zonation beat correlates with magnetic this unit and furthermore to be indistinguishable polarity chrons M22-M25. However, this part of from the Glance Conglomerate [Kluth, 1982; the magnetic polarity sequence is now considered Vedder, 1984]. This pole is referred to here as to be entirely Oxfordian in age or approximately the Glance Conglomerate pole rather than the 156-163 Ma [Harland et al., 1982]. The correla­ Canelo Hilla pole to avoid confusion with recent tion proposed by Steiner is based largely on the paleomagnetic data from the Canelo Hilla vol­ dominance of reversed polarity observed in the canics aensu strictu [May, 1985]. Morrison rather than on detailed matching of A stable primary component of magnetization polarity patterns. was isolated in 15 sites through af and thermal An alternative age assignment which is more cleaning techniques. Both a positive reversals consistent with both magnetostratigraphic and test and a conglomerate test indicate that the paleontologic data is that the Morrison Formation isolated retnanence is primary. The age of the is dominantly Tithonian to late Kimmeridgian in ash flow tuf f s in the lower Glance Conglomerate age and that the polarity zonation is best is well constrained by a Rb/Sr iaochron date of correlated with chrons Ml6 to Ml9 [May, 1985]. 151+2 Ma [Kluth et al., 1982]. Fossil mammals from the Brushy Basin Member of Steiner [1983] expresses some reservation the Morrison Formation in Colorado and Wyoming regarding the use of the Glance Conglomerate have many taxa in common with the lower Purbeck (Canelo Hills) pole because these rocks are "in Beds (late Tithonian) in England [Clemens et al., the tectonically disturbed area, southern basin 1979], and Colbert [1973] has discussed the and range of Arizona." She states that "no inde­ similarity in dinosaur faunas between Morrison, pendent evidence exists to prove a lack of rota­ Purbeck, and Tendaguru (Africa) localities. The tions subsequent to magnetization." Kluth et al. latter occurrence is considered late Kimmeridgian [1982] address this question and point to a or early Tithonian on the basis of ammonites. number of reasons why vertical axis rotation 'ts Steiner [1983] shows the age of the Morrison unlikely. Paleomagnetic sites were distributed Formation as mid-Oxfordian to mid-Tithonian in two sections with different attitudes and citing Imlay [1980], but she also states that the separated by numerous faults yet simple struc­ uppermost 25 samples from the Norwood locality tural correction significantly improves were actually collected from the overlying Burro clustering. Pre-basin and range paleogeographic Canyon Formation which is generally considered to reconstructions restoring about 203 crustal be Early Cretaceous in age [Tschudy et al., extension do not require vertical axis rotation 1984]. The cleaned directions from these 25 [Coney, 1978]. May et al. [this issue] discuss samples are indistinguishable from those of the the various forms of structural orientation data upper Morrison Formation (Brushy Basin Member), and paleomagnetic data which indicate that no suggesting close temporal proximity. Imlay'a significant vertical axis tectonic rotation took [1980] belief that the Morrison may be as old as place in southeastern Arizona during the Laramide middle Oxfordian appears to be based on the fact or Basin and Range orogenies. Perhaps most that the underlying Sundance Formation contains importantly, the Corral Canyon pole [May et al., fossils only as young as early Oxfordian. How­ this issue] from the Patagonia Mountains is con­ ever, Pipiringos and O'Sullivan [1978] recognize sistent spatially and temporally with the Glance a principal unconformity (JS) between the Conglomerate pole even though these two areas are Morrison and Sundance formations which exhibits separated by a major Laramide structure as much as 46 m of relief over a distance of 6 km (Lampshire Canyon-Dove Canyon fault, see Kluth near Escalante, Utah. Thus considerable time [1982]). Similarly, both poles are consistent in could separate deposition of the Sundance and terms of age and location with both older and Morrison formations at least on a local scale. younger reliable paleopoles from other parts of Therefore, on the basis of vertebrate fossils the craton. and a revised magnetostratigraphic correlation, the Morrison Formation is probably late Middle Jurassic1 163-188 Ma Kimmeridgian through Tithonian in age or approxi­ mately 152-144 Ma. Using the alternative Two paleopolea are considered to be reliable magnetoatratigraphic correlation of May [1985], indicators of the Middle Jurassic paleofield for we have assigned absolute ages of 145 and 149 Ma North America. The 171+3 Ma Corral Canyon pole to the upper and lower Morrison poles, respec­ from southeastern Arizona [May et al., this tively. These ages are somewhat younger than issue] and the 179+3 Ma Newark Trend Group II those assigned by Gordon et al. [ 1984] (147-152 pole provide reference poles for Bathonian and and 152-156 Ma) and still may err in being some­ Bajocian time respectively (Figure 2). No reli­ what too old. While Harland et al. [1982] pick able paleopoles are known from Callovian or the Kimmeridgian-Tithonian boundary at 150 Ma, Aalenian age rocks in North America. their chronogram for this interval is poorly Corral Canyon. A paleomagnetic pole from the constrained and includes no data for 142-148 Ma. Corral Canyon rocks in the Patagonia Mountains of This stage/age boundary may therefore be somewhat southeastern Arizona has been discussed by May et younger, as suggested by Van Hinte [1976], and al. [this issue]. This pole was based on 11 the Morrison poles may be as young as 136-141 Ma. sites within welded ash flow tuffs and a single Glance Conglomerate. Kluth et al. [1982] site in red mudatone all of which were shown to report a paleomagnetic pole from Late Jurassic carry stable magnetizations by extensive af and ash flow tuff s in southeastern Arizona (Table 1). thermal demagnetization. The data satisfy all Although originally considered to be a "lower reliability criteria, and the age is well con­ member" of the Canelo Hills Volcanics, these strained isotopically as 172±5.8 Ma. The rocks have recently been shown to rest deposi- Patagonia Mountains pole can be criticized on the May and Butler1 Jurassic Apparent Polar Wander 11 ,539 same grounds as the Glance Conglomerate pole (see frequent polarity reversals. Paleomagnetic above), and the same arguments in defense of its results from late Bathonian equivalent volcanic tectonic stability can be invoked. rocks and red mudstones in the Patagonia Newark Trend Group II. A plethora of paleo­ Mountains, SE Arizona [May et al., this issue], magnetic results have been published from Early do not exhibit anomalous NRM intensities, retain and Middle Jurassic igneous rocks of the Newark stable primary magnetizations, and are dqminantly trend intrusive series. Rather than discuss each of normal polarity through 650 m of section. It of these poles separately, many of which are VGPs seems likely that the complicated magnetizations rather than true paleopoles, the reader is reported for Jurassic sedimentary rocks from the referred to the comprehensive summary of Smith western interior United States reflect sedimento­ and Noltimier [1979]. These workers recognized logic and diagenetic processes rather than geo­ that VGPs from the Newark Trend intrusives magnetic field behavior. cluster into two groups that correspond to Summerville Formation. Nearly all recent temporally distinct intrusive episodes. The analyses of Jurassic APW have included a refer­ older (Pliensbachian equivalent) Group I or "pre­ ence pole from the Middle Jurassic Summerville folding" dikes and sills yield a mean pole posi­ Formation published by Steiner [1978]. In tion at 63.0°N, 83.2°E (Figure 2) based on data eastern Utah the Summerville consists of approxi­ 0 from 72 sites with k•56 and a 95=2.J • The Group mately 120 m of thin bedded red siltstone and II or "postfolding" intrusives yield a pole at fine-grained sandstone and is overlain uncon­ 65.3°N, 103.2°E (Figure 2) based on 156 sites formably by the Morrison Formation. The age of with k•92 and a 95 .. 1.4°. the Summerville is considered to be middle to The ages of these two intrusive groups are early late Callovian by Pipiringos and O'Sullivan well constrained by 39Ar/40Ar and 40K/t+OAr radio­ [1978] and late Callovian by Imlay [1980] or metric data reported by Sutter and Smith [1979]. approximately 163-167 Ma. Their work suggests ages of 179+3 Ma and 195+4 Ma The Summerville paleomagnetic results suffer for the Group II and Group I pOi.es, respectively. from a strong Cenozoic normal polarity overprint, Some controversy exists regarding the age range although Steiner [1978] concludes that much of of various extrusive volcanic units within the the section has a reversed polarity primary com­ Newark Supergroup sequence and included within ponent. Thermal demagnetization to 6300-660oc the Group I average pole. Although the Watchung was considered to have isolated a stable primary basalts, the Granby tuff and the Holyoke basalt magnetization, but in only 15 out of 391 samples yield VGPs interpreted as belonging to the Group collected. Steiner [1978] argues that because I cluster by Smith and Noltimie~ [1979], both this set of normal and reversed directions is fossil fish and fossil pollen from surrounding antipodal, their mean is a "good estimate" of the sediments indicate an age of Hettangian to Summerville direction but admits that the Summer­ Sinemurian (200-213 Ma). Using VGPs from the ville pole is "approximate at best." Because the West Rock, Mt. Carmel, and East Rock intrusives unambiguous isolation of a stable primary compo­ only, Smith and Noltimier report a paleopole nent has not been demonstrated and the number of samples is very low, the pole position from the located at 63.1 °N, 82.5oE (k=l07, a 9 =2.8°) which is nearly identical with the Group t pole indi­ Summerville Formation is considered unreliable. cating that the inclusion of data from the extru­ Twin Creek Formation. McCabe et al. [1982] sive rocks does not significantly affect the published a paleomagnetic pole from the Middle mean. The Watchung basalts have been recently Jurassic Twin Creek Formation in Wyoming. Seven restudied by Mcintosh et al. [1985], who conclude of 10 sites considered stable by McCabe et al. that the number of flows sampled was probably [1982] are located within the Prospect and Darby insufficient to average secular variation even thrust sheets of the Wyoming overthrust belt. though the associated k value was only 26. Paleomagnetic data from the Chugwater Group Interestingly, the mean pole position calculated (Early Triassic) in these same thrust sheets were for the Watchung basalts is intermediate between cited by Grubbs and Van der Voo [1976] as evi­ the Group I and Group II poles of Smith and dence for differential tectonic rotation of Noltimier [1979] perhaps suggesting that the frontal thrusts associated with impingement on older lavas were partially or wholly remagnetized basement cored uplifts of the foreland. during regional heating associated with the Early McCabe et al. [1982] acknowledge these earlier and Middle Jurassic phases of dike and sill results but argue that because the mean direc­ emplacement. The radiometric age of the Group II tions observed for three sites from the Twin pole is correlative with the Bathonian-Bajocian Creek on the Gros Ventre foreland block are boundary of Harland et al. [1982]. statistically indistinguishable from site direc­ Steiner [1978, 1980, 1983] has investigated tions in the thrust sheets, the formation mean the paleomagnetism of other Middle Jurassic pole is representative of cratonic North America. formations in northern Arizona and northern Of these three foreland sites, however, two have Wyoming. These include the Bajocian Gypsum k values of less than 15 and 95 greater than 20° Spring and Piper formations, the Bathonian to leaving only a single site that passes the early Callovian Carmel and Rierdon formations, acceptance criteria described above. Also, and the late Callovian to Oxfordian Swift deletion of these two sites decreases the total Formation. No reliable Jurassic paleomagnetic number of sites to eight, less than the 10 site poles have been obtained from any of these rocks acceptance criterion. In addition, McCabe et al. because of ubiquitous and dominant secondary [1982] state that the anomalous easterly location magnetizations attributed by Steiner to postdepo­ of the Twin Creek pole may be the result of sitional overprint. Steiner [1980] suggests that postdepositional chemical remagnetization. low NRM intensity and complex multicomponent Stump Formation. A Jurassic paleomagnetic magnetization also may reflect Jurassic geomag­ pole from the Wyoming overthrust belt published netic field properties of low intensity and/or by Schwartz and Van der Voo [1984] for the 11 ,540 May and Butler: Jurassic Apparent Polar Wander

Oxfordian Stump Formation is not included within calculated from the mean of site (=sample) VGPs. our set of reliable poles for reasons similar to Resorting to polarity interval mean poles as those discussed above for the Twin Creek Forma­ VGPs, a formation mean pole based on N=7 is t ion. Most importantly, only seven sites w.ere located at 62.1°N, 10.2°E (Cl 95=6.3°, k•92.2). studied, and the between site k value is less The confidence interval on this calculation is than 20. Although the Stump magnetization seems probably larger than it should be, The Cl95 to pass a fold test, the overlying Early Creta­ values for poles cited by Steiner and Helsley are ceous sediments also studied by Schwartz and Van 6.8° for polarity zones (N•7) and 2.5° for der Voo [1984) show clear evidence for remag­ samples (N=l05); the weighted standard error netization during folding. confidence parameter used by Gordon et al. [1984) White Mountain Magma Series. Often used as a is 4°. reliable 180 Ma paleopole for cratonic North A disturbing feature of the Kayenta data set America, the White Mountain magma series pole of is the noncircular distribution of polarity zone Opdyke and Wensink [1966) is considered unreli­ VGPs. These poles are elongate along the path of able. The White Mountain pole is indistinguish­ Early and Middle Jurassic APW, from about the able from the geographic North pole in contrast position of the Wingate Formation pole (i.e., to all other reliable Early through Middle Sinemurian) almost to the Newark Trend Group II Jurassic poles which fall along a band of present pole 079+3 Ma). Steiner and Helsley [1974) latitude at 60°-65° N. The distribution of VGPs suggested-a great circle fit to these poles from the 12 stable sites is markedly streaked approximately 90° away from the sampling locality ranging from 71.5°N, 46.0°E to 7 5.5°N, 188.5°E. which they attributed to possible long-term In a general sense, this VGP streak mimics the variation of the geomagnetic field. Within-zone Early Jurassic-Early Cretaceous APW path but is directions are not obviously streaked (except for displaced into higher latitudes, suggesting both zone N2), but there is a clear difference between a protracted and complex Mesozoic magnetization VGPs calculated from normal polarity zones and history as well as incorrect structural those calculated from reversed zones. A pole corrections or present field overprints. calculated from the four reversed-zone VGPs at Furthermore, there are systematic directional 0 61.4°N, 63.0°E (a95=5.o , k=357.l) is identical differences between individual intrusions, as to the paleopole from the underlying Wingate pointed out by Steiner and Helsley [1972). Formation. The mean pole of the three normal The intrusive history of the White Mountains polarity zone VGPs lies farther east, although it magma series is known to be complex with radio­ is not statistically distinct (62,3°N, 80,loE, metric dates ranging from 235 to 100 Ma [Foland a 95 =17.6°, k=50). These features of the Kayenta and Faul, 1977]. K-Ar dates for the intrusions Formation paleomagnetism are puzzling and deserve sampled by Opdyke and Wensink range from 180 to further investigation. However, we do consider 118 Ma, although no thermal demagnetization the Kayenta pole of Steiner and Helsley to be a results were reported by these workers. Five of reliable Early Jurassic reference. the 12 stable sites were collected from intru­ Wingate Formation. A pole from the Wingate sions with K-Ar dates of 118 and 121 Ma, while Formation has been discussed by Steiner [ 1983] only three sites were from the White Mountains and Gordon et al. [1984) based on original data pluton for which dates of 168-180 Ma have been of Reeve [1975). Steiner [1983) concluded that a obtained by Foland and Faul [1977). Until a reliable paleopole could not be calculated from systematic pluton-by-pluton paleomagnetic study Reeve's data because of extensive present field of the White Mountains magma series incorporating overprint that was not completely removed. detailed thermal demagnetization is conducted, Gordon et al. [1984) concluded differently and this pole cannot be considered a reliable calculated a pole position based on 156 samples Jurassic reference pole for North America. from two localities. They describe a small reversed overprint removed by thermal demagneti­ Early Jurassic: 188-213 Ma zation at 550°-630°c. After having reviewed Reeve's thesis we are inclined to agree with the Three reliable paleopoles are known from Early conclusion of Gordon et al. [1984); however, the Jurassic rocks and one from rocks whose estimated locality mean k values reported by Reeve are age includes the Triassic-Jurassic boundary. The quite low. The Wingate Formation pole is tenta­ 195+4 Ma Newark Trend Group I pole discussed tively included as the earliest reliable Jurassic above, the Kayenta Formation pole, and the reference for North America (Table 1, Figure 2). Wingate Formation pole record APW during Peterson and Pipiringos [1979) and Imlay [1980) Pliensbachian and Sinemurian time (Figure 2). both consider the Wingate to be Sinemurian in age Kayenta Formation. A paleomagnetic pole from or approximately 200-206 Ma. the Kayenta Formation was reported by Steiner and Manicouagan Impact Structure. The youngest Helsley [1974). At that time, the Kayenta was reliable Triassic paleopole for North America is thought to be upper Triassic, but Olsen et al. from igneous rocks of the Manicouagan Impac~ [1982) and Imlay [1980) have shown it to be structure in Quebec, Canada, whose radiometric Pliensbachian in age, or approximately 194-200 age (215±4 Ma) includes the Triassic-Jurassic Ma. As discussed in relation to the Morrison boundary. This date is based on concordant whole Formation, the sampling scheme of Steiner and rock and mineral separate K/Ar data published by Helsley is not especially well suited for paleo­ Wolfe [1971) (revised for new decay and abundance magnetic pole position calculation because constants). A combination of paleomagnetic "sites" are represented by single cores closely results based on the work of Robertson [1967) and spaced throughout a stratigraphic interval or by Larochelle and Currie [1967) yields a pole posi­ polarity intervals. tion at 58.8°N, 89.9°E ( a95=5.8°) (Figure 2). Data listed in Table 1 of Steiner and Helsley Paleomagnetic properties of a variety of rock [1974) do not allow a paleomagnetic pole to be types were studied with both af and thermal May and Butlers Jurassic Apparent Polar Wander 11,541 demagnetization. We have averaged VGPs from six diabase from Pennsylvanias Further results, sites from Robertson [1967] with the five site .:!:. Geophys. Res., 1.1_, 5673-5687, 1972. mean VGPs of Larochelle and Currie [1967]. Beck, M.E., Jr., D-Yscordant paleomagnetic pole Irving and Irving [ 1982] list 10 equally positions as evidence of regional shear in weighted poles from northeastern North America fthe western Cordillera of North America, Am • igneous rocks of supposed Late Triassic and Early .:!:. Sci., 276, 694-712, 1976. - Jurassic age. All of these poles are now con­ Beck, M.E., Jr., Paleomagnetic record of plate sidered to belong to either the Group I or Group margin tectonic processes along the western II Newark Trend poles of Smith and Noltimier edge of North America, .:!:. Geophys. Res., 85, [1979] or to be unreliable VGPs. The postfolding 7115-7131, 1980. intrusions pole (49) is essentially the Group II Beck, M.E., Jr., On the mechanism of tectonic pole of Smith and Noltimier [1979] which is the transport in zones of oblique subduction, reference cited by Irving and Irving [1982], but Tectonophysics, 93, 1-11, 1983. they misquote the age as 170 Ma rather than 179 Beck, M.E., Jr., Introduction to the special Ma and they cite the pole at 68.4°N, 98.9°E issue on correlations between plate motions rather than at 65.3°N, 103.2°E as given by Smith and Cordilleran tectonics, Tectonics, 3, 103- and Noltimier. The prefolding intrusions pole 105, 1984. - (53) of Irving and Irving is basically identical Beck, M.E., Jr., and C.D. Burr, Paleomagnetism to the Group I pole of Smith and Noltimier. and tectonic significance of the Goble The Anticosti Island diabase dike pole (50) is Volcanic Series, southwestern Washington, clearly a VGP only and should not be given equal Geology, l• 175-179, 1979. weighting. Larochelle [1971] stated that the Bogen, N.L., D.V. Kent, and R.A. Schweickert, mean pole based on 11 cores from two sites Carboniferous Maroon and upper ments of the Passaic Formation. A recent rein­ Permian-lower Triassic State Bridge Formations vestigation of the paleomagnetism of the Passaic in north central Colorado, M.S. thesis, Univ. Formation concludes that it carries an ''unremov­ of Tex. at Dallas, Richardson, 1974. able" secondary magnetization of presumed Ceno­ Clemens, W.A., J.A. Lillegraven, E.H. Lindsay, zoic age [Mcintosh et al., 1985]. Site mean and G.G. Simpson, Where, when, and what--A directions fail the fold test and k values for survey of known Mesozoic mammal distribution, normal and reverse polarity sites are extremely in Mesozoic Mammalss The First Two-Thirds of low (8 and 6). Mammalian History, edited by J.A. Lillegraven, The "Connecticut Volcanic rocks," "Diabase z. Kielan-Jaworowska, and W.A. Clemens, pp. 7- dikes and sills," "Newark Diabase," and "North 58, University of California Press, Berkeley, Mountain Basalt" poles of Irving and Irving Calif., 1979. [1982] also are based on data included by Smith Colbert, E.H., Continental drift and the distri­ and Noltimier [1979] in their Group I and Group bution of fossil reptiles, Implications of II poles. Continental Drift to the Earth Sciences, edited by D.H. Tarling, and S.K. Runcorn, NATO Acknowledgments. We thank Greg Cole for Adv. Stud. Inst. Sec., Ser. £, .!• 395-412, supplying the computer program used in the PEP 1973. analysis. The authors benefited from numerous Collinson, D.W., and S.K. Runcorn, Polar discussions of this work with Myrl Beck, Peter wandering and continental drifts Evidence of Coney, and Bill Dickinson and from a helpful paleomagnetic observations in the United review by Jack Hillhouse. This research was States, Geol. Soc. Am. Bull., l!• 915-958, funded by National Science Foundation grant EAR- 1960. 8314113. Coney, P.J., The plate tectonic setting of southeastern Arizona, Land of Cochise, edited by J.F. Callender, J.C. Wilt, and R.E. References Clemons, Fieldl Conf. Guideb • .!!:.. M. Geol. Soc., 29, 285-290, 1978. Baag, c., and C.E. Helsley, Evidence for penecon­ Coney, P.J., Accretionary tectonics in western temporaneous magnetization of the Moenkopi North America, Relations of Tectonics to Ore Formation, .:!:. Geophys. Res., .z!, 3308-3320, Deposits in the Southern Cordillera, edited by 1974. W.R. Dickinson and W.D. Payne, Ariz. Geol. Beck, M.E., Jr., Paleomagnetism of Upper Triassic Soc. B.!&:..o 14, 23-38, 1981. ---- 11 ,542 May and Butler: Jurassic Apparent Polar Wander

Coney, P.J., and S.J. Reynolds, Cordilleran Gordon, R.G., A. Cox, and s. O'Hare, Benioff zones, Nature, 270, 403-406, 1977. Paleomagnetic Euler poles and the apparent Coney, P.J., D.L. Jones, and J.W.H. Monger, polar wander and absolute motion of North Cordilleran suspect terranes, Nature, 288, America since the Carboniferous, Tectonics, 1• 329-333, 1980. 499-537. 1984. Cordell, L., Extension in the Rio Grande rift, J. Gradstein, F.M., and R.E. Sheridan, On the Geophys. Res., 87, 8561-8569, 1982. - Jurassic Atlantic Ocean and a synthesis of Cox, A., Latitude dependence of the angular dis­ results of Deep Sea Drilling Project leg 76, persion of the geomagnetic field, Geophys. J. Initial Rep. Deep Sea Drill. Proj., 76, 913- R. Astron. Soc., 20, 253-269, 1970. 943, 1983. Dalrymple, G.~ c.s. Gromme, and R.W. White, Grubbs, K.L., and R. Van cier Voo, Structural Potassium-argon age of diabase dikes in deformation of the Idaho-Wyoming overthrust Liberia: Initiation of central Atlantic belt, as determined by Triassic paleomag­ rifting, Geol. Soc. Am. Bull., 86, 399-4ll, netism, Tectonophysics, 33, 321-336, 1976. 1975. Hamilton, w., Plate-tectonic mechanism of Damon, P.E., M. Shafiqullah, and K.F. Clark, Age Laramide deformation, Contrib. Geol., 19, 87- trends of igneous activity in relation to 92, 1981. metallogenesis in the southern Cordillera, Harland, W.B., A.V. Cox, P.G. Llewellyn, C.A.G. Relations of Tectonics to Ore Deposits in the Pickton, A.G. Smith, and R. Walters, ! Southern Cordillera, edited by W.R. Dickinson Geologic :!'.!!!.!. Scale, Cambridge University and W.D. Payne, Ariz. Geol. Soc. ~ 14, Press, New York, 1982. 137-154, 1981. Harrison, C.G.A., and T. Lindh, A polar Demarest, H.H., Jr., Error analysis for the wandering curve for North America during the determination of tectonic rotation from paleo­ Mesozoic and Cenozoic, :I!_ Geophys. Res., 87, magne tic data, .:!:.. Geophys. Res., 88, 4321- 1903-1920, 1982. 4328, 1983. Herrero-Bervera, E., and C.E. Helsley, Paleomag­ deBoer, J., Paleomagnetic-tectonic study of Meso­ netism of a polarity transition in the lower zoic dike swarms in the Appalachians, J. (?)Triassic Chugwater Formation, Wyoming, .:!:.. Geophys. Res., ll• 2237-2250, 1967. Geophys. Res., 88, 3506-3522, 1983. deBoer, J., Paleomagnetic differentiation and Hillhouse, J.W., Paleomagnetism of the Triassic correlation of the Late Triassic volcanic Nikolai Greenstone, McCarthy Quadrangle, rocks of the central Appalachians (with Alaska, Can • .:!:.. Earth Sci., 14, 2578-2592, special reference to the Connecticut Valley), 1977. Geol. Soc. Am. Bull., 79, 609-626, 1968. Hillhouse, J.W., and C.S. Gromme, Paleomagnetism DuBois, P.M., E. Irving, N.D. Opdyke, S.K. of the Triassic Hound Island Volcanics, Runcorn, and M.R. Banks, The geomagnetic field Alexander terrane, southeastern Alaska, J. in upper Triassic times in the United States, Geophys. Res., 85, 2594-2602, 1980. - Nature, 180, ll86-ll87, 1957. Hillhouse, J.W., and c.s. Gromme, Northward dis­ Eby, G.N., Geochronology of the Monteregian Hills placement and accretion of Wrangellia: New alkaline igneous province, Quebec, Geology, paleomagnetic evidence from Alaska, J. ~. 468-470, 1984. Geophys. Res., 89, 4461-4477, 1984. - Eisbacher, G.H., Evolution of successor basins in Hillhouse, J.W., C.S. Gromme, and T.L. Vallier, the Canadian Cordillera, Modern and Ancient Paleomagnetism and Mesozoic tectonics of the Geosynclinal Sedimentation, edited by R.H. Seven Devils volcanic arc in northeastern Dott, and R.H. Shaver, Spec. Puhl. Soc. Econ. Oregon,.:!:.. Geophys. Res., 87, 3777-3794, 1982. Paleontol. Mineral., .!!• 274-291, 1974. Imlay, R.W., Correlation of the Jurassic forma­ Engebretson, D.C., Relative motions between tions of North America exclusive of Canada, oceanic and continental plates of the Pacific Geol. Soc. Am. Bull., 63, 953-992, 1952. Basin, Ph.D. dissertation, Stanford Univ., Imlay, R.W., Jurassic paleobiogeography of the Stanford, California, 1982. coterminous United States in its continental Fisher, R.A., Dispersion on a sphere, Proc. R. setting, U.S. Geol. Surv. Prof. ~ 1062, Soc. London, Ser. A, 217, 295-305, 195~ - 67-99, 1980. Foland, K.A., and H-; Faul, Ages of the White Irving, E., Paleomagnetism and its Application to Mountain intrusives - New Hampshire, Vermont, Geological and Geophysical Problems, 399 pp., and Maine, USA, Am • .:!:.. Sci., 277, 888-904, John Wiley, New York, 1964. 1977. Irving, E., Drift of the major continental blocks Forsyth, D.W., and s. Uyeda, On the relative since the Devonian, Nature, 270, 304-309, importance of the driving forces of plate 1977. - motion, Geophys. :I!_!:.. Astron. Soc., 43, 163- Irving, E., Paleopoles and paleolatitudes of 200, 1975. North America and speculations about displaced Francheteau, J., and J.G. Sclater, Paleomagnetism terrains, Can • .:!:.. Earth Sci., _!!, 669-694, of the southern continents and plate 1979. tectonics, Earth Planet. Sci. Lett., 6, 93- Irving, E., and G.A. Irving, Apparent polar 106, 1969. ------wander paths Carboniferous through Cenozoic Ghosh, N., S.A. Hall, and J.F. Casey, Seafloor and the assembly of Gondwana, Geophys. Surv., spreading magnetic anomalies in the Venezuelan 5, 141-188, 1982. Basin, The Caribbean-South American Plate Irving, E., and Park, J.K., Hairpins and super­ Boundary and Regional Tectonics, edited by intervals, Can • .:!:.. Earth Sci., .2_, 1318-1324, W.E. Bonini, R.B. Hargraves, and R. Shagam, 1972. Mem. 162, Geol. Soc. Am., 65-80, 1984. Irving, E., J.W.H. Monger, and R.W. Yole, New May and Butlers Jurassic Apparent Polar Wander 11 ,543

paleomagnetic evidence for displaced terranes Paleomagnetism and oxide mineralogy of upper in British Columbia, The continental crust and Triassic red beds and basalts in the Newark its mineral deposits, edited by P.W. Basin, Geol. ~ Am. Bull., 96, 463-480, Strangway, Spec. Paper. Geol. Assoc. Can., 20, 1985. 441-456, 1980. Merrill, R.T., and M.W. McElhinny, .!!!.!. Earth's Irving, E., G.J. Woodsworth, P.J. Wynne, and A. Magnetic Field: Its History, Origin, and Morrison, Paleomagnetic evidence for displace­ Planetary Perspective, Int. Geophys. Ser., ment from the south of the Coast Plutonic vol. 32, 401 pp., Academic, Orlando, Fla., Complex, British Columbia, Can. J. Earth Sci., 1983. 22, 584-598, 1985. Monger, J.W.H., and E. Irving, Northward dis­ Jones, D.L., N.J. Silberling, and J.W. Hillhouse, placement of north-central British Columbia, Wrangellia: A displaced terrane in north­ Nature, 285, 289-294, 1980. western North America, Can • .:!:_ Earth Sci., 14, Monger, J.W.H., R.A. Price, and D.J. Tempelman­ 2565-2577, 1977. Kluit, Tectonic accretion and the origin of Klitgord, K.D., P. Popenoe, and H. Schouten, the two major metamorphic and plutonic welts Florida1 A Jurassic transform plate boundary, in the Canadian Cordillera, Geology, 10, 70- .:!:_ Geophys. Res., 89, 7753-7772, 1984. 75, 1982. Kluth, C.F., Geology and mid-Mesozoic tectonics Moore, G.F., and D.E. Karig, Structural geology of the northern Canelo Hills, Santa Cruz of Nias Island, Indonesia: Implications for County, Arizona, Ph.D. dissertation, Univ. of subduction zone tectonics, Amer • .:!:_ Sci., 280, Ariz., Tucson, 1982. 193-223, 1980. Kluth, C.F., R.F. Butler, L.E. Harding, M. Morgan. W.J., Hotspot tracks and the early Shafiqullah, and P.E. Damon, Paleomagnetism of rifting of the Atlantic, Tectonophysics, 94, late Jurassic rocks in the northern Canelo 123-139, 1983. Hills, southeastern Arizona, .:!:_ Geophys. Res., Oldow, J.S., H.G. Ave'Lallemant, and W.J. 87. 7079-7086. 1982. Schmidt, Kinematics of plate convergence Larochelle, A., Note on the paleomagnetism of two deduced from Mesozoic structures in the diabase dykes, Anticosti Island, Quebec, Proc. western Cordillera, Tectonics, 1• 201-227, Geol. Assoc. Can., 23, 73-76, 1971. 1984. Larochelle, A., and K.L. Currie, Paleomagnetic Olsen, P.E., A.R. McCune, and K.S. Thompson, study of igneous rocks from the Manicouagan Correlation of the early Mesozoic Newark Structure, Quebec, .:!:_ Geophys. Res., 72, 4163- Supergroup by vertebrates, principally fishes, 4169, 1967. Am.:!..:_ Sci., 282, 1-44, 1982. Mankinen, E.A., Paleomagnetic evidence for a Late Opdyke, N.D., The paleomagnetism of the New Cretaceous deformation of the Great Valley Jersey Triassic1 A field study of the inclina­ sequence, Sacramento Valley, California, :!..:_ tion error in red sediments, .:!:_ Geophys. Res., .!!.!:_U.S. Geol. Surv., !• 383-390, 1978. 66, 1941-1949, 1961. Manspeizer, W., Early Mesozoic basins of central Opdyke, N.D., and H. Wensink, Paleomagnetism of Atlantic passive margins, in Geology of rocks from the White Mountain plutonic­ Passive Continental Margins: History, ~ volcanic series in New Hamshire and Vermont, ture, and Sedimentologic Record (With Special :!..:_ Geophys. Res., ll• 3045-3051, 1966. Emphasis .!?.!!. the Atlantic Margi~dited by Palmer, A.R., The decade of North American geol­ A.W. Bally, Educ. Course Note Ser., vol. 19, ogy 1983 geologic time scale, Geology, .!!,, pp. 4-1 - 4-60, American Association of 503-504, 1983. Petroleum Geologists, Tulsa, Okla., 1981. Peterson, F., and G.N. Pipiringos, Stratigraphic May, S.R., Paleomagnetism of Jurassic volcanic relations of Navajo Sandstone to Middle rocks in southeastern Arizona and North Jurassic formations, southern Utah and American Jurassic apparent polar wander, Ph.D. northern Arizona, U.S. Geol. Surv. Prof. ~ dissertation, Univ. of Ariz., Tucson, 1985. 1035-B, 43 pp., 197~------May, S.R., R.F. Butler, M. Shafiqullah, and P.E. Pindell, J. L., Alleghenian reconstruction and Damon, Paleomagnetism of Jurassic volcanic subsequent evolution of the Gulf of Mexico, rocks in the Patagonia Mountains, southeastern Bahamas, and Proto-Caribbean tectonics 1 Arizonai Implications for the North American Tectonics, 4, 1-40, 1985. 170 Ma reference pole,:!..:_ Geophys. Res., this Pipiringos, G:N., and R.B. O'Sullivan, Principal issue. unconformities in Triassic and Jurassic rocks, May, S.R., P.J. Coney, and M.E. Beck, Jr., Paleo­ western interior United States - a preliminary magnetism and suspect terranes of the North survey, U.S. Geol. Surv. Prof. ~ 1035-A, American Cordillera, U.S. Geol. Surv. Open 29 pp., 1978. File~ 83-799, 8 pp., 1983. Rees, c. J., E. Irving, and R.L. Brown, Secondary McCabe, c., R. Van der Voo, and B.H. Wilkinson, magnetization of Triassic-Jurassic volcani­ Paleomagnetic and rock magnetic results from clastic rocks of the Quesnel terrane, Quesnel the Twin Creek Formation (Middle Jurassic), Lake, B.C., Geophys. Res. Lett., g, 498-501, Wyoming, Earth Planet. Sci. Lett., 60, 140- 1985. 146, 1982. Reeve, S.C., Paleomagnetic studies of sedimentary McElhinny, M.W., Paleomagnetism and Plate rocks of Cambrian and Triassic age, Ph.D. Tectonics, 358 pp., Cambridge University dissertation, Univ. of Tex. at Dallas, Press, New York, 1973. Richardson, 1975. McElhinny, M.W., and R.T. Merrill, Geomagnetic Reeve, S.C., and C.E Helsley, Magnetic reversal secular variation over the past 5 m.y., Rev. sequence in the upper portion of the Chinle Geophys., 13, 687-708, 1975. Formation, Montoya, New Mexico, Geol. Soc. Am. Mcintosh, W.C., R.B. Hargraves, and C.L. West, Bull., 83, 3795-3812, 1972. ------11 ,544 May and Butlers Jurassic Apparent Polar Wander

Robertson, W.A., Manicouagan, Quebec, paleomag­ Sundvik, M., R.L. Larson, and R.S. Detrick, netic results, Can. :!..:_ Earth Sci., i• 641- Rough-smooth basement boundary in the western 649, 1967. North Atlantic basins Evidence for a Saleeby, J.B., Ocean floor accretion and seafloor-spreading origin, Geology, _!!, 31-34, volcanoplutonic arc evolution of the Mesozoic 1984. Sierra Nevada, in The Geotectonic Development Sutter, J.F., and T.E. Smith, 40Ar/39Ar ages of of California, edited by W.G. Ernst, pp. 132- diabase intrusions from Newark trend basins in Iil, Prentice-Hall, Englewood Cliffs, N. J., Connecticut and Maryland: Initiation of 1981. central Atlantic rifting, Am. :!..:_ Sci., 279, Saleeby, J.B., G.C. Harper, A.W. Snoke, and W.D. 801-831, 1979. Sharp, Time relations and structural­ Symons, D.T.A., Paleomagnetism of the Jurassic stratigraphic patterns in ophiolite accretion, Island Intrusions of Vancouver Island, British west central Klamath Mountains, California, J. Columbia, Can. :!..:_ Earth Sci., !. 1388-1396, Geophys. Res., 87, 3831-3848, 1982. 1970. Schouten, H., and K.D. Klitgord, The memory of Symons, D.T.A., New paleomagnetic data for the the accreting plate boundary and the continu­ Triassic Guichon batholith, southcentral ity of fracture zones, Earth Planet. Sci. Britiah Columbia, and their bearing on Terrane Lett., 59, 255-266, 1982. --- -- I tectonics, Can. :!..:_ Earth Sci., 20, 1340- Schwartz, S.Y., and R. Van der Voo, Paleomagnetic 1344, 1983. study of thrust sheet rotation during foreland Symons, D.T.A., and C.R. Litalien, Paleomagnetism impingement in the Wyoming-Idaho overthrust of the lower Jurassic Copper Mountain Intru­ belt, :!..:_ Geophys. Res., 89, 10,077-10,086, sions and the geotectonics of Terrane I, 1984. British Columbia, Geophys. !!..!:.. Lett., .!!• Shive, P.N., and W.E. Frerichs, Paleomagnetism of 685-688, 1984. the Niobrara Formation in Wyoming, Colorado, Tschudy, R.H., B.D. Tschudy, and L.C. Craig, and Kansas, :!..:_ Geophys. Res., 1.2.• 3001-3007, Palynological evaluation of Cedar Mountain and 1974. Burro Canyon formations, Colorado Plateau, Shive, P.N., M.B. Steiner, and D.T. Huycke, U.S. Geol Surv. Prof. !!.P.:.• 1281, 24 pp., Magnetostratigraphy, paleomagnetism, and 1984. remanence acquisition in the Triassic Van Alstine, D.R., Apparent polar wandering with Chugwater Formation of Wyoming, :!..:_ Geophys. respect to North America since the late Res., 89, 1801-1815, 1984. Precambrian, Ph.D. dissertation, Calif. Inst. Smith, T.E., Paleomagnetic study of lower Meso­ of Technol., Pasadena, 1979. zoic diabase dikes and sills of Connecticut Van Alstine, D.R., and J. deBoer, A new technique and Maryland, Can.:!..:_ Earth Sci., g, 597-609, for constructing apparent polar wander paths 1976. and the revised Phanerozoic path for North Smith, T.E. and H.C. Noltimier, Paleomagnetism America, Geology, !• 137-139, 1978. of the Newark trend igneous rocks of the north Van Eysinga, F.W.B. (compiler), Geological Time­ central Appalachians and the opening of the table, 3rd ed., Elseview, Amsterdam, 1975. central Atlantic Ocean, Am • .b_ Sci., 279, 778- Van Hinte, J.E., A Jurassic time scale, Am. 807, 1979. ABBoc. Pet. Geol. Bull., 60, 489-497, 1976.-- Steiner, M.B., Mesozoic apparent polar wander and Vedder, L.K., Stratigraphic relationship of Late Atlantic plate tectonics, Nature, 254, Jurassic Canelo Hills Volcanics and Glance 107-109, 1975. Conglomerates, southeastern Arizona, M.S. Steiner, M.B., Magnetic polarity during the thesis, Univ. of Ariz., Tucson, 1984. Middle Jurassic as recorded in the Summerville Vilas, J.F,A., Paleomagnetism of South American and Curtis Formations, Earth Planet. Sci. rocks and the dynamic proceBSes related with Lett., 38, 331-345, 1978. --- -- the fragmentation of western Gondwana, in Steiner, M.B., Investigation of the geomagnetic Paleoreconstruction of the Continents, Geodyn. field polarity during the Jurassic, J. Ser., vol. 2, edited by M.W. McElhinny and Geophys. Res., !5, 3572-3586, 1980. - D.A. Valencio, PP• 106-114, AGU/GSA, Steiner, M.B., Mesozoic apparent polar wander and Washington, D. C., 1981. plate motions of North America, in Mesozoic Wolfe, S.H., Potassium-argon ages of the Paleogeography of West-Central United States, Manicouagan-Mushalagan Lake structure, :!..:.. edited by M.W. Reynolds and E.D. Dolly, pp. 1- Geophys. Res., ll• 5424-5436, 1971. 11, Rocky Mountain Section, Society of Yole, R.W., and E. Irving, Displacement of Economic Paleontologists and Mineralogists, Vancouver Islands Paleomagnetic evidence from Denver, Colo., 1983. the Karmutsen Formation, Can. J. Earth Sci., Steiner, M.B., Is the Colorado Plateau rotated?, .!1.· 1210-1288, 1980. ------Eos Trans. AGU, 65, 864, 1984. Steiner, M.B., and""'G.E. Helsley, Jurassic polar movement relative to North America, J. R. F. Butler, Department of Geosciences, Geophys. Res., J.]_, 4981-4993, 1972. - University of Arizona, Tucson, AZ 85721. Steiner, M.B., and C.E. Helsley, Magnetic polar­ s. R. May, Exxon Production Research Co., P.O. ity sequence of the Upper Triassic Kayenta Box 2189, Houston, TX 77242. Formation, Geology, ~. 191-194, 1974. Steiner, M.B., and C.E. Helsley, Reversal pattern and apparent polar wander for the Late (Received January 16, 1986; Jurassic, Geol. Soc. Am. Bull., 36, 1537-1543, revised June 19, 1986; 1975. ------accepted June 24, 1986.)