““3"" ’0 v~ I'.'1_m’ "Hm ‘ ', "-l‘ " ._.. -) - . ' A RATIONALE FOR THEORIGINS or“ ~ ‘ ‘ :‘ . = MASSIF ANORTHOSITES’V ’ Dissertation for the Degree of Ph} D. MICHIGAN STATE UNIVERSITY GRAHAM RYDER ' 1974 f - This is toicertify that the -' thesis entitled ARationale for the Origins of Massif Anorthosi'bes presented by Graham Ryder has been accepted towards fulfillment of the requirements for Ph, D. degree in Geology flew Major professor Date August 8, 1974 0-7639 ABSTRACT A RATIONALE FOR THE ORIGINS OF MASSIF ANORTHOSITES BY Graham Ryder The origins of massif anorthosites cannot be simply explained by a single magma type, and two of the commonly proposed parents for anorthosites are andesites (quartz- diorites) and high-alumina basalts. This thesis proposes that these two magmas are the parents for two separate groups of anorthosites which include all massifs, and that the parents for any given anorthosite massif can be determined by the rock sequence associated with that massif. Evidence from experiments and from phenocrysts in volcanics suggests that andesites crystallizing in the granulite facies would produce plagioclase cumulates at the base (anorthosites), followed by dioritic and acidic material, whereas high-alumina basalts would produce gabbros followed by anorthosite with very little succeeding acidic material. All massif anorthosites for which relevant data are available have one or the other of these strati- graphic sequences. Grouped according to these sequences 0 050 \Q Graham Ryder they coincide with two previously proposed groups, i.e., Andesine-type and Labradorite-type, whose characteristics are shown to be compatible with derivation from andesite and high-alumina basalt, respectively. Two anorthosites which demonstrate the vertical sequence of rocks expected from the fractional crystalliza- tion and gravity settling of an andesite magma were studied as a test of the model. The San Gabriel anorthosite suite in California demonstrates a sequence in which the plagio- clase composition, determined optically and with the micro- probe, becomes more sodic from the anorthosite (mean approximately An48) through diorite (s An ) to monzonite 38 and quartz-monzonite (z An22)' Cryptic stratigraphy is present in the major lithological units and there is no hiatus in the cryptic sequence, suggesting that the sequence is comagmatic. The San Gabriel anorthosite suite therefore has not only a lithological but also a cryptic stratigraphy very similar to that expected from the fractional crystal- lization of an andesite magma. The Langelier anorthosite suite in Quebec has a border zone sheared from the anorthosite proper, which therefore is probably diapiric. Although the measured plagioclase compositions indicate that the sequence is compatible with an andesite parent, there must be some doubt as to the validity of the results because of the diapiric nature of the pluton which puts the relations of the lithologies present in doubt. Graham Ryder The evidence suggests that andesitic and high- alumina basaltic magmas give rise to two independent groups of anorthosites. Because these two magmas are characteristic of present-day subduction zones, it is possible that anorthosite massifs are indicative of paleo- orogeny and paleo-subduction. A RATIONALE FOR THE ORIGINS OF MASSIF ANORTHOSITES BY Graham Ryder A DISSERTATION Submitted to Michigan State University in partial fulfillment of the requirements for the degree of DOCTOR OF PHILOSOPHY Department of Geology 1974 This thesis is dedicated to COLLEEN M. ROBICHAUD . . . Every man of science whose outlook is truly scientific is ready to admit that what passes for scientific knowledge at the moment is sure to require correction with the progress of discovery; nevertheless it is near enough to the truth to serve for most practical purposes, though not for all. In science, where alone something approxi- mating to true knowledge is to be found, mens' attitude is tentative and full of doubt. - Bertrand Russell "Ying tong iddle i poo." - Spike Milligan ii ACKNOWLEDGMENTS My sincere thanks go: to Tom Vogel, for assistance, not only on a professional but also on a personal level, which enabled this work to be started, continued, and completed when times would otherwise have been rough. Without his help, nothing would have been done; to Robert Ehrlich for introducing me to the dif- ference between science and natural history, and for assistance and encouragement not only in this but in all other work in which I have been involved; to my committee, Dr. C. Spooner, Dr. H. Bennett, and Dr. H. Stonehouse, especially for their indulgence and patience during the later stages of all this. Especial thanks are expressed to Dr. Spooner who aided in many ways in work which helped to clarify my thoughts; to Bob Malcuit who went with me to California and did most of the work; to Rich Wharton and Wendy Shaft who went with me to Quebec, and especially to Wendy for doing the plagioclase determinations on the Quebec samples; to all my friends, whether members of the cliquish, conceited, all-talk-no-action group or not, and especially iii to Bruce, Gary, Kevin, Lee, and Steve, who all made my stay at MSU enjoyable; to my family, for missing me, for waiting, for being patient, and for being encouraging; to the Geological Society of America, for providing funds for the field work in the form of a Penrose Bequest Fund Research Grant; and finally, but not least, to Colleen, for making me think about the way I am. iv TABLE OF CONTENTS LIST OF TABLES . . . . . . . . . . LIST OF FIGURES . . . . . . . . . . INTRODUCTION 0 O O O O O O O O O O MASSIF-TYPE ANORTHOSITES--PRELIMINARY DISCUSSION THE CRYSTALLIZATION OF BASALTIC AND HIGH-ALUMINA BASALTIC MAGMAS IN THE GRANULITE FACIES . . THE CRYSTALLIZATION OF ANDESITE MAGMA IN THE GRANULITE FACIES . . . . . . . . . EXAMPLES OF THE TWO DIFFERENT SEQUENCES . . TWO GROUPS OF MASSIF ANORTHOSITE . . . . POSSIBLE TESTS OF THE MODEL . . . . . . FIELD TESTING OF THE MODEL: APPROACH . . . THE SAN GABRIEL ANORTHOSITE SUITE, CALIFORNIA ANALYTICAL METHODS FOR PLAGIOCLASE COMPOSITIONS RESULTS AND DISCUSSION . . . . . . . . THE LANGELIER ANORTHOSITE, QUEBEC . . . CONCLUSIONS AND SPECULATIONS . . . . . . LIST OF REFERENCES 0 O C O O O O O Page -vi Vii 10 12 18 25 28 30 41 43 54 S7 60 Table 1. LIST OF TABLES Anorthosite massifs grouped according to their vertical sequence of rock types, i.e., stratigraphy . . . . . . . . Comparison of the optical and microprobe data . . . . . . . . . . . . Optically derived plagioclase compositions Microprobe-derived plagioclase compositions vi Page 20 42 44 46 Figure. 1. 10. 11. 12. l3. 14. LIST OF FIGURES Sequences expected from the crystallization of an andesite and a high-alumina basalt in the granulite facies . . . . . . . Sequence at the Bell River Complex . . . . CaO-MgO-FeO-Alkalis diagram (FeOt + Fe203 + FeO) to show the compatibility of the Honey Brook Anorthosite with an andesite magma O O O O O I O O I O O O O Enstatite/Anorthite ratios, adapted from Anderson and Morin (1969) . . . . . . San Gabriel Anorthosite: Geological map, adapted mainly from Carter and Silver (1972) o o o o o o o o o o o 0 San Gabriel Anorthosite: Sample localities . San Gabriel Anorthosite: Locality map . . Photographs of rock types in the San Gabriel snite O O O O O O O O O I O 0 Photograph of the inhomogeneous diabase at sample sites 63-67 . . . . . . . . San Gabriel Anorthosite: Histogram of plagioclase compositions . . . . . . San Gabriel Anorthosite: Contour map of plagioclase compositions . . . . . . Microprobe trace of a mesoperthite from sample 35 (monzonite) . . . . . . . Postulated reconstruction of the structure of the central part of the San Gabriel massif . . . . . . . . . . . . Generalized map of the Langelier Anorthosite. vii Page 15 13 24 31 34 35 38 39 47 48 49 52 55 INTRODUCTION Anorthosites and gabbroic anorthosites, rocks composed predominantly of plagioclase, are found in three main geological environments: (1) large massif-type plutons; (2) layers, boudins and fault slices associated with aluminous metamorphic rocks; (3) layers in stratified basic intrusions (Anderson, 1969). This thesis is con— cerned with the origins of massif-type anorthosite. Although plutonic massif-type anorthosite is well- known and widespread, its origin is problematical. Anderson and Morin (1969) have divided this anorthosite type into two groups based on lithological and mineralogi- cal characteristics, and proposed that one type crystallized from a basic magma; the other was produced by remelting of the first group by a world-wide and probably catastrophic reheating event. The groups seem to have been accepted (e.g., Wynne-Edwards, 1972), but the model has not been widely accepted. There are indeed obvious flaws in the model: the catastrophic event could not have been synchronous and experimental evidence in general indicates that melting of the first group could not have produced the second. A world-wide heating event of the necessary intensity would have remelted much of the Earth's crust, l and this is not evident in the geological record--even the parent group plutons show no evidence at the present time of ever having undergone remelting. This thesis proposes that a more generally accept- able model based on realistic and realizable magmas is desirable, and considers the possibility that each pluton of massif-type anorthosite is derived by the fractional crystallization of either an andesitic magma or a high- alumina basaltic magma under granulite metamorphic conditions. MASSIF-TYPE ANORTHOSITES-- PRELIMINARY DISCUSSION Massif-type anorthosites are found only in high- grade metamorphic terrains, usually of the granulite facies (Anderson, 1969). An attempt to explain the limited age variation of anorthosite (essentially Protero- zoic), a preoccupation of many authors, is therefore inseparable from an attempt to explain the limited age variation in ages derived from the "mobile belts" (Anheusser et_al., 1969) in which they are found. The ubiquitous associated presence of pyroxene granulite assemblages with anorthosite cannot be ignored. Heat flow from the Earth is unlikely to have been substantially greater one b.y. ago than it is at present (Lubimova, 1969), and the regional metamorphism is probably taking place today, for instance, deep in the crust in island arcs (Miyashiro, 1972). Thus the limited appearance of the granulite facies, and that of anorthosites, in the geo- logical record since the Archean is most likely due to the level of erosion, and anorthosite-producing regimes may well exist at depth at the present time. Although massif-type anorthosites are common (Anderson, 1969), anorthosite lavas have never been 3 reported. Experimental evidence lends little support to the possibility of the production of primary anorthosite magmas by partial melting of mantle or crustal materials. For instance, the Yoder and Tilley (1962) method of pro- ducing an anorthosite magma by the melting of basalt will in fact produce a magma approximately andesitic in compo- sition (Holloway and Burnham, 1972). For these reasons, as well as the unreasonably high liquidus temperature of such a magma (Luth and Simmons, 1969; Isachsen, 1969), the possibility of magmas of anorthosite composition must be rejected. Most workers therefore agree that anorthosite is produced from the accumulation of plagioclase gravitation-j ally either by floating or settling from a non-anorthositic magma. Many authors have favored a gabbroic anorthosite magma, generally charged with plagioclase crystals in sus- pension (Buddington, 1939, 1969; Bowen, 1971), or at least some form of high-alumina basalt (Emslie, 1965). Others have preferred to postulate an intermediate magma (Balk, 1931; Philpotts, 1966). These two magmas have generally been postulated for specific massifs, and then generalized by the author to all_massifs. Some plutons currently included in the category of massif anorthosite can be adequately explained as being a crystallization differentiate from a variant of high- alumina basaltic magma, and any radically different magma would not easily explain their chemical and mineralogical nature and variations. Such massifs include the Michikamau, Labrador, intrusion (Emslie, 1965) which apparently has a chilled margin of high-alumina basalt, and the Kadavur, India, massif (Subramaniam, 1956). On the other hand, other massifs are extremely difficult to explain as being derived from a basaltic magma, mainly because of their high sodium contents and their lack of ultramafic materials, the latter being a necessary part of a differentiated basaltic sequence. These massifs include the Roseland, Virginia, massif (Herz, 1969a) and the Allard Lake, Quebec, massif (Hargraves, 1962). Much of the problem concerning the parent magma of the massif anorthosites stems from the fact that two distinct anorthosite types have been traditionally treated as one group for which consequently no single magma type is compatible with all the evidence. For example, it is invalid to use evidence from the Michikamau to explain the evolution of the Marcy (Adirondacks) massif (Buddington, 1969). A genetic model for a specific massif must be built upon evidence obtained only from that massif until it becomes evident that other massifs are so petrologically and chemically similar that they must have a similar petrogenesis. All massif anorthosites can, however, be produced from some form of high-alumina basalt or some form of andesite. It is the purpose of this thesis to evaluate 6 from experimental data what might be expected from the crystallization of the two magma types in the granulite facies of regional metamorphism, especially in terms of the sequences of mineral assemblages which would be evolved during fractional crystallization. This data is then com- pared with the petrological characteristics of anorthosite massifs. The first approach is to obtain such comparisons from data already available in the literature. The second is to use the model to make detailed predictions about two anorthosite massifs which, according to the model, have an andesitic parent and to evaluate, through field sampling and laboratory analysis, the comparison with the predictions. The assumption is made here that plagioclase, and all other minerals, will sink and not float in the magma. Morse (1973) has discussed this problem and pointed out that even in a theoretically favorable instance for plagioclase flotation (Kiglapait), the plagioclase in fact settled. While by no means proven, the assumption is felt to be justified. THE CRYSTALLIZATION OF BASALTIC AND HIGH-ALUMINA BASALTIC MAGMAS IN THE GRANULITE FACIES Fractional crystallization trends of tholeiitic and alkali basaltic magmas are fairly well documented in terms of the field evidence (Wager and Brown, 1967). Plagioclase has not been found to develop at an early stage in plutonic sequences, and olivine is usually the earliest phase, succeeded by oxides and pyroxenes. The dominant phenocryst phase in basaltic volcanics is also olivine, indicating that it is the liquidus phase at the depth of first crystallization. Abundant experimental evidence from both melting and crystallization of basaltic magmas (Yoder and Tilley, 1962; Green and Ringwood, 1964, 1967; Holloway and Burnham, 1972; and others) shows that plagio- clase is absent from the liquidus over a wide variety of pressures, water contents, and basaltic compositions. Only at very low pressures, approaching atmospheric, and with low water contents, approaching zero, can plagioclase be produced at a basaltic liquidus. At pressures above minimal but below about ten kilobars olivine is the liquidus phase; above this pressure it may be replaced by orthopyroxene (Green and Ringwood, 1964). 7 Experimental evidence indicates that in a high- alumina basalt, at least when the partial pressure of water is low, plagioclase is produced at or near to the liquidus between zero and 6.8 kilobars, but it is usually accompanied by both pyroxene and olivine (Green, Green, and Ringwood, 1967). Analogies with other compositions ranging from basalt to andesite indicate that the presence of water depresses the plagioclase crystallization temperature to below that of the mafic phases, and thus completely removes plagioclase from the liquidus of a high-alumina basalt. With such experimental evidence as a base, it is possible to predict the course of fractional crystallization of a basaltic or a high-alumina basaltic magma in the granulite facies. Assuming that gravity settling is operative a basaltic magma will produce a vertical sequence with ultramafites at the base, with succeeding gabbro, some interlayered anorthosite, and diorite, followed by minor amounts of iron-rich acidic rocks. A high-alumina basalt under the same conditions and with a low water content will produce a sequence with minor basal ultramafites, succeeded by gabbro, much more gabbroic anorthosite, some diorite, and with minor acid differentiates at the top (Figure 1). In both cases the plagioclase becomes more sodic as the sequence evolved. .mowomm ouHHssmum on» cw Auomwnv panama nowadamunmfio m can Aumoav ouwmoosm no mo sowumeHHmummHo mop Eoum omuoomxm moosmsvmmII.H musmflm 3:92.22: 2... .. lit/xii! 352.0 0.: 030... 0 136: .0332... .: o was. 95.3: 609.0 a 3009093 b:.......:.:.. no. a 6030 «I'll f 09. 0 PE. . + “0222.0th . : Ac t .r ... ++ v ONOZ «.095 >30: 0 OoU r. 08 16 Green's (1969) crystallization of a synthetic andesite at 9 kilobars, with that of the Honey Brook (from Crawford et_§l., 1971), and with that of the Skaergaard intrusion, calculated from Wagner and Brown (1967). Green's synthetic andesite trend is extrapolated beyond the experimental data on the basis of the plots of both the parent liquid and the residual liquid. The early stage of the experimental trend shows a deflection towards magnesium compared to the Honey Brook trend and compared to the parent liquid. This is due to the difficulty of analys- ing the iron-bearing phases and the trend must curve back towards and around the parent liquid, away from the M90 apex according to simple mass balance principles, during the later stages. It should be noted that at 13.5 kilobars the early part of the trend is further from the starting liquid than it is at 9 kilobars; therefore presumably it is closer to the starting liquid at 6 to 7 kilobars, the expected pressures of the granulite facies, and even closer at lower pressures which conceivably apply to the crystal- lization of the Honey Brook anorthosite suite. The Honey Brook trend is quite close to the trend demonstrated for the synthetic andesite, and if the plagio- clase compositions of the rocks can be taken as an indicator of the direction of crystallization, then the direction is also appropriate. This trend is valid for about 50 per cent of crystallization of the synthetic liquid; the l7 remainder at Honey Brook might well be represented by the quartz-monzonite, and if so then the curving trend is even closer to that of the synthetic andesite. The trend for the Skaergaard basaltic cumulates is very different. Although superficially the trends in the fields of CaO-MgO-FeOt and CaO-FeOt-Alkalis appear similar to the other trends, the rocks being compared at similar stages are very different: peridotites at Skaergaard com- pared to anorthosites at Honey Brook and in the synthetic products. It is therefore unlikely that the Honey Brook anorthosite suite represents crystallization from a basaltic magma. The synthetic andesite magma experiment showed that the liquidus plagioclase at 6 kilobars would have a compo- sition of about An and the maximum anorthite content of 55’ plagioclase at Honey Brook is of this composition. The sequence of rocks at Honey Brook is therefore very close in its characteristics to the sequence expected from the crystallization of an andesite magma at pressures equivalent to high grades of metamorphism. The quartz-monzonite at Honey Brook is likely to be consanguineous, an hypothesis which might be tested by geochemical methods. TWO GROUPS OF MASSIF ANORTHOSITE If massif anorthosites do in fact fall into two groups, one derived from an andesitic magma and the other derived from a high-alumina basaltic magma, then their vertical sequences, as outlined in the previous sections, should be distinctive. It should then be possible to divide massif anorthosites into two groups using the two different sequences as discriminators, and a given massif should fit easily into one or the other of the groups even if only a part of the sequence is visible. This assumes in practice that the vertical sequence is derivable from and available for a given massif. The two groups, if they exist, can then be further investigated to see if their features are in fact compatible with the proposed magmas. There is a lack of information concerning the internal variability and compositional variations within many anorthosite massifs, and many bodies, such as the Matamec (Greig, 1945), may be too deformed to allow objective assessments of their pre-tectonic vertical sequences. However, the sequences for 16 massifs have been gleaned from the literature. Nine massifs have sequences similar to that expected from an andesite magma crystallizing in the granulite facies, while the other seven 18 19 massifs have sequences as expected from a high-alumina basalt under similar conditions (Figure 1). Thus positive grouping based on igneous stratigraphy does occur. The massifs are listed in Table 1. Two more plutons, including the Matamec, may have andesitic sequences, and are included in Table 1. Anderson and Morin (1969) divided massif anortho- sites into two groups, on criteria not including the vertical sequence. Their division is based essentially on the plagioclase compositions, the predominant rock types present, the types of ferromagnesians, and the types of oxide minerals. Their Labradorite-type consists mainly of gabbroic anorthosite, and contains labradorite, clino- pyroxene, olivine, and titaniferous magnetite. Their Andesine-type is characterized by more true anorthosite, by andesine, by orthopyroxene with no olivine, and by hemo- ilmenite. The andesine-type is generally domical while the Labradorite-type is frequently irregularly deformed. The plagioclase of a massif anorthosite derived from an andesitic magma should be more sodic than the plagio- clase in a gabbroic anorthosite derived from a more basic high-alumina basaltic magma. Table 1, showing plagioclase compositions as listed in the literature, indicates that the "andesitic" group has mainly andesine, and the "high- alumina basaltic" group has mainly labradorite. This immediately suggests that there may be some correlation 20 TABLE l.--Anorthosite massifs grouped according to their vertical sequence of rock types, i.e., stratigraphy. Massif Type* P12922¢i25e Reference .NMESTEKISHJENCE: Adirondacks (A) 46-52 Davis, 1969; De'Waard, 1970 Allard Lake (A) 40-52 Hargraves, 1962 Bethoulat 55 Neale, 1965 BjerkreimrSogndal - Michot & Michot, 1969 Honey Brook (A) 38-55 Crawford et_§1,, 1971 Labrieville (A) 35-53 Anderson, 1966 Langelier - Rondot, 1961 Morin (A) 47-56 Philpotts, 1966 St. Urbain (A) 45 Mawdsley, 1927 2Matamec (A) 45-50 Greig, 1945 ?Wilkinson (A) 45-53 Harrison, 1944 HKHFAUTHNA SEQflflEE: Angola (L) 50-80 Simpson and Otto, 1960 . Bell River (L) 60-65 Freeman, 1939; Freeman and Black, 1944 Kadavur (L) 50-55 Subramanian, 1956 Iofoten (L) 50-60 Raney, 1969, 1971 Michikamau (L) 54-72 Emslie, 1969 Nain (L) 34-58 Wheeler, 1960 .Vfital (L) 53 Blais, 1960 * A = Andesine-type; L = Labradorite-type (Anderson & Morin, 1969). 21 between these groups and those of Anderson and Morin (1969). Anderson and Morin list 10 plutons as being of Labradorite- type and 15 as being Andesine-type. The relationship between their classification and the classification proposed here, based on a petrogenetic model, and using predicted vertical sequences as discriminators, is shown in Table 1. It is evident that the two independent sets of criteria lead to identical groupings, as all "andesitic" group plutons which are also listed by Anderson and Morin are included in their Andesine-type and vice versa. Similarly all "high-alumina basaltic" group plutons which are also listed by Anderson and Morin are included in their Labradorite type and glee yergg. No mis-classifications occur. (The Bell River and Angola complexes satisfy the criteria for Labradorite-type and have been included as such.) Since the two methods of classification lead to identical groupings, the conclusion must be that two fundamentally different types of massif anorthosite exist. The groups can be distinguished not only on vertical sequence, but also plagioclase composition, the types of ferromagnesians and oxides, and dominant rock types. Labradorite-type anorthosites are mineralogically similar to basaltic rocks, and many authors have postulated basic magmas as the parent of individual plutons which belong to this group. Anderson and Morin (1969), in their discus- sion of the group, relate Labradorite—type anorthosites to a 22 basaltic parent. The ubiquitous olivine may in particular be evidence of basicity. The Michikamau massif seems to be a clear example of a basic pluton in that it has a chilled margin of basalt, and Emslie (1965, 1969) has proposed a basaltic parent magma. The lack of associated acid derivatives would be expected from a basic magma. The Nain massif, with its associated suite of adamellitic rocks (Wheeler, 1960) seems to be an exception, but there is evidence (below) that the adamellite was a later intrusion, and not comagmatic with the anorthositic rocks. Morse (1972) has proposed, on petrological grounds, that the parent of the Nain anorthosite was a high-alumina basalt, and that the adamellitic rocks originated in a separate and later event. A number of authors have proposed intermediate magmas for specific Andesine-type anorthosites. Balk (1931) prOposed a dioritic parent for the Adirondacks suite, Oakeshott (1937) pr0posed a similar parent for the San Gabriel massif, and more recently Philpotts (1966) has proposed an acid dioritic parent for the Morin Series in Quebec. Overall bulk compositions for anorthosite-syenite suites are estimated to be generally dioritic with no unusual chemical characteristics, e.g., Buddington (1939, p. 235), Barth (1936, p. 301) and Philpotts (1966, p. 51), and therefore are compatible with an andesitic magma as the parent. 23 The plagioclase of Andesine-type anorthosite is generally andesine (Table l) and is also typically anti- perthetic. The orthoclase mol per cent is frequently of the order of 6 to 8. This is compatible with derivation from an andesite magma which contains about 10 to 11 per cent of orthoclase constituents (McBirney, 1969), but is not compatible with derivation from a basaltic magma. Compared with layered basaltic plutons, anortho- sites have a high mol per cent enstatite in orthopyroxene/ mol per cent anorthite in plagioclase ratio (Anderson, 1969). No adequate theory exists for this feature (Anderson and Morin, 1969), but Green's (1969) anorthositic materials produced from a synthetic andesite has composi- tions and ratios similar to those of Andesine-type anortho- sites (Figure 4). Empirically, therefore, the feature is compatible with an andesitic parent for the group. Similar data on experimental high-alumina basalt does not exist. While the precise nature of the parental magmas of the massif anorthosites must remain a matter of conjec- ture with the available descriptive information, there appear to be no petrological obstacles to pr0posing a high- alumina basaltic magma for the Labradorite-type massifs, and to proposing an andesitic parent for the Andesine-type massifs. 24 .Amomav owuoz cam acmuoosfl Scum owommom .mofiumu wuwsuuocfl\muwumumomll.w wusmflm 0.? Om 0.9 mm :4 3 .05. . w m. up 3 U x .— x m < x < o._. 0 0 mlooOo ‘~0.H 0. O“ 00’ O on. 000 00 I 0 o .2 0 ~00. 0.. 2. U. 0 O O O . D. O 35 Inca .530“ . .mmE huflamooq “ouflmosuuocd Howuoow ommll.h ousmam N ”EX 0 i as“ 2915852 £02502 ; . com to O>COU Ubt cwmuo “2.33 a» S. . 502302 0602 . 30m t0>cou 0o 36 l. Anorthosites--plagioclase. 2. Diorites (Jotunites*)--p1agioclase, pyroxenes largely altered to hornblende, mesoperthites. 3. Monzonites**(Syenites*)--mesoperthites, pyroxenes largely altered to hornblende, minor plagioclase, quartz. 4. U1tramafites—-meta111c ores, apatite, olivine (?) 5. Hornblende Gabbro—-hornb1ende, bytownite. 6. Diabase--bytownite, labradorite, pyroxenes, hornblende, olivine, biotite. representative figures of these rock types are shown in Figure 8, A to F. Groups 5 and 6 were omitted from the analysis because they do not appear to be intimately related to the other rocks in the massif. Both of these groups are extremely fresh compared to the anorthosites, diorites and monzonites. The the anorthosite zone at Mount in the eastern border zone in and may itself be of Mesozoic the diorites on the southwest hornblende gabbro intrudes Gleason, at Rabbit Peak, and contact with Mesozoic granites, age. The diabase intrudes side at Sand Canyon and on the Santa Clara Truck trail as a broad inhomogeneous (Figure 9) dyke-like formation and has a very fresh recrystallized mineralogy. The direct relation of these * Terminology of Carter and Silver, 1972. ** Microprobe scan shows that the plagioclase in the meSOperthite is An 20, and thus the rock is more appropriately termed a monzonite. 37 0 Figure 8.--Photographs of rock types in the San Gabriel 'suite: . A - Anorthosite - Diorite Monzonite - Ultramafite IF.) U 0 U1 l - Hornblende gabbro Diabase '11 l 38 A. Crossed Polarizers B. 1mm i533, . D. Plane Polarized' lmm E. Outcrop F. Crossed Polarizers 1mm 39 .hmlmm moufim onEmm pm wmmomflo msomsomOEoocfl osu mo snowmouonmll.m wuomflh 40 two groups to the anorthosite suite is unlikely. Group 4, the ultramafites are of a rock type commonly associated with anorthosite. They are assumed to be comagmatic with the anorthosite suite, probably separating during an immiscible stage, as has been suggested by Anderson (1966) for those at the Labrieville anorthosite. They have no bearing on the present study because of their lack of plagioclase. ANALYTICAL METHODS FOR‘ PLAGIOCLASE COMPOSITIONS Plagioclase compositions were determined on crushed samples using the revised Tsuboi Dispersion Method (Morse, 1968), because it is both rapid and precise (Morse, 1968), and because of its superior accuracy over other optical methods (Vogel, unpublished data). Approximately 20 grains per sample were determined and the mean value accepted. Precision in this study was such that the standard error for a sample was equivalent to less than 2% Anorthite, and usually about 1% Anorthite. Microprobe analyses were made for two reasons. The first was to check the accuracy of the optical determina- tions, and the second was to obtain compositions for the small grains of rare primary plagioclase in the monzonite, a task beyond the capability of the optical method. All microprobe determinations were made using an Applied Research Laboratories EMX instrument. The analyses were made using an accelerating potential of 14.5 Kv. and with a sample current of 0.02 uamps. Albite, Orthoclase, and Anorthite (An 95) were used as standards, and corrections were made using the method of Smith and Ribbe (1966). Because the data required in this study was essentially 41 42 anorthite mol per cent, the corrected data was recalculated to 100% (Or + Ab + An). The microprobe results for samples also determined optically are shown in Table 2, and demon- strate that the optical data is consistently 2-3% An greater than the microprobe determinations. For the com- parative purposes of this study, this difference can lead to no significant error. TABLE 2.--Comparison of the optical and microprobe data. Optical Microprobe Sample Mean An% Range* Mean An% Range* SG-36 22 21%-22% 21 19-23 56-44 11 10-12 9 4-14 SG-78 47% 46-49 45 43-47 SG-79 50 49-51 47% 45%-49% SG-134 33 31-35 30% 28%-32% * As one standard error on each Side of the mean. RESULTS AND DISCUSS ION The optical data are displayed in Table 3, and the complete micrOprobe data in Tables 2 and 4. The data is also shown in histogram form in Figure 10, and in a hand- contoured form in Figure 11. A trace across a typical mesoperthite from group 3 is shown as Figure 12, and demon- strates that the rock is apprOpriately termed a monzonite rather than a syenite because its feldspar is at least 50% plagioclase containing more than 5% anorthite. The results demonstrate conclusively that the plagioclase in the anorthosite suite varies continuously from being more calcic in the anorthosite, to less calcic in the diorites, to even less calcic in the monzonites. This suggests that the crystallization sequence was anorthosite, followed by diorite, and finally by monzonite. The variation is continuous rather than disjunct (Figure 10), suggesting that differentiation from a common magma is likely. The sizes of the individual modes are insig- nificant except insofar as they represent the sampling and therefore the outcrop area, and there is no certainty that the outcr0p area represents the relative volumes of the rock types differentiated from the common magma. The diabase and the hornblende gabbro both plot separate from 43 44 TABLE 3.--Optically derived plagioclase compositions (see text). Sample An% Rock Sample ' An% Rock 1 43 A 46 51 A 2 51% D 47 41% A 3 51 A 48 44% A 4 52 A 49 41 A 5 52 A 50 40 A 6 50 A 51 43 A 7 52 A 52 49% A 8 49% A 53 44 A 9 44 D 54 53% A 10 50% D 55 49 A ll 0 A 56 53 A 12 44 A 57 - H 13 49 A 58 75 H 14 28 D 59 43 A 15 31 D 60 46 A 16 27% D 61 47 A 17 25% D 62 46 A 18 46% A 63 - W 19 47 A 64 - W 20 50 A 65 - W 21 47 A 66 75 W 22 47 A 67 - W 23 45 A 68 41 D 24 49 A 69 39% D 25 46 A 70 35 D 26 50 A 71 35 A 27 42% A 72 - M 28 43% A 73 40 A 29 50 A 74 39 A 30 40 A 75 41 A 31 41 A 76 42 A 32 — M 77 45 A 33 - M 78 47% A 34 - M 79 50 A 35 - M 80 49% A 36 22 M 81 41 A 37 - M 82 46% A 38 - M 83 49 A 39 - A 84 - Albitite 40 - U 85 49 A 41 - M 86 50 A 42 - M 87 49 A 43 - M 88 - M 44 11 M 89 - M 45 - M 90 29 D TABLE 3.--Continued. 45 Sample An% Rock Sample An% Rock 91 31% D 130 - U 92 35 D 131 - W 93 35% D 132 36 D 94 - D 133 36% D 95 38 D 134 33 D 96 40 D 135 38 D 97 42 D 136 - D? 98 40% D 137 - U 99 43 D 138 - ? 100 - D 139 39 D 101 52 A 140 - ? 102 51 A 141 49 A 103 75 H 142 - ? 104 47% A-D 143 49 A 105 53% A 144 - ? 106 53% A 145 49% A 107 54 A 146 54 A 108 46% A 147 55 A 109 45% A 148 - A 110 34 D 149 55% A 111 49 A 150 49 A 112 - M 141 49% A 113 - M 152 - M 114 48 A 153 - M 115 47% A 154 - M 116 ‘36 A 155 - M? 117 - M 156 26 D 118 35% D 157 41 A 119 - M 158 - A 120 - M 159 30 D 121 — M 160 - ? 122 - M 161 31 D-A 123 50 A 162 33% A+Quartz 124 40 A 163 43% A 125 47% A 164 47% A 126 39 A 165 40 A 127 - W 166 45 A 128 - @ 167 42% A 129 - W 168 41% A A - Anorthosite/Leuconorite D - Diorite M - Monzonite or Quartz-monzonite U - Ultramafite H - Hornblende Gabbro W - Diabase 46' TABLE 4.--Microprobe-derived plagioclase compositions (see text). Sample Grain An% 35 1 20.9 36 1 21.3 36 2 22.0 36 3 21.0 38 1 22.5 38 2 21.0 43 1 27.1 43 2 23.6 44 1 10.1 44 2 8.7 44 3 8.4 78 1 44.6 78 2 45.5 79 1 45.9 79 2 49.0 79 3 48.0 88 1 27.0 88 2 20.3 122 1 20.3 134 1 29.2 134 2 30.4 134 3 30.5 47 .mGOHuwmomfioo ommHUOAmmam mo Emnooumwm “Hofluoow smmll.oa ousmflm wm