The Deccan Traps, India is composed of lava flows with enormous volumes and the duration of volcanism has been of great interest relating to the mechanism of their formation. Further, the age of the Deccan Traps has called the attention from the point of view that the age of the Deccan Traps volcanism might have been almost simultaneous with that of the Cretaceous-Tertiary Boundary(KTB) which has been connected with the major extinction event. In order to reveal the detailed volcanic history of the Deccan volcanism, 40Ar-39Ar datings were performed to twenty-seven lavas and nine dykes from the Deccan Traps. In addition, three biotite samples which were separated from alkalic rocks were dated by the newly developed laser heating 40Ar-39Ar dating method. Since the 40Ar-39Ar method accompanied with the stepwise heating has merits to obtain more reliable ages compared to the K-Ar method, it was applied to the Deccan Traps samples which were often affected by the excess Ar and/or Ar loss. Further, to evaluate the reliability of the obtained 40Ar-39Ar age, the relationship between the 40Ar-39Ar age pattern and the nature of sample has been examined in detail in this thesis. Some criteria to evaluate the obtained ages have been determined and applied to the results.
In order to accomplish this purpose, improvement of the 40Ar-39Ar dating system and the development of the laser heating 40Ar-39Ar dating system were performed. By the combination of a sector-type mass spectrometer, VG-3600 and a radio-frequency induction heating system, analytical procedures of the 40Ar-39Ar dating have been established to obtain precise and accurate 40Ar-39Ar ages. The laser heating 40Ar-39Ar dating system has also been developed to obtain the 40Ar-39Ar age from a sample with small amount, whose reliability was confirmed by measuring the age standard samples, MMhb-1, LP-6 and so on. The laser heating 40Ar-39Ar dating method was applied to three biotite samples from the alkalic rocks of the Deccan Traps and it has given more reliable ages than those of whole rock samples. To improve the accuracy of the obtained age data, the age of intralaboratory age standard sample(EB-1) was redetermined precisely as 91.4 +/- 0.5 Ma which was intercalibrated by the international age standard, LP-6. The age of EB-1 has been used as the basis for 40Ar-39Ar dating experiments in this thesis.
Although it is possible to evaluate the reliability of 40Ar-39Ar dating results by itself to some extent, the criteria are not always clear and their interpretations sometimes differ among different investigators. Hence, in order to establish the criteria to evaluate the 40Ar-39Ar dating result, the relationship between the 40Ar-39Ar dating result and the nature of the sample has been examined. Based on the data distribution in the inverse isochron plots of the 40Ar-39Ar dating results of the Deccan Trap samples, I have grouped them into typical five types. Samples of Type 1 show a good isochron with the atmospheric (36Ar/40Ar) for the trapped component. A good plateau 40Ar-39Ar age is obtained for samples of this type. This type is found from fresh and secondarily undisturbed samples. Samples of Type 2 also show a relatively good isochron, but trapped (36Ar/40Ar) components are smaller than that of the air. Saddle shape pattern is observed in the 40Ar-39Ar age spectrum for samples of Type 2. The presence of excess Ar is expected in samples of Type 2. This type is often found from dyke samples. Samples of Type 3 show the data distribution of the lower temperature fractions in the upper-right region to the reference isochron in the inverse isochron plot. In the K/Ca plot and age spectrum, samples of this type indicate the characteristics of Ar loss from them. Samples of this type generally contain some amounts of altered materials. Samples of Type 4 show the data distribution of an ellipse shape around the reference isochron in the inverse isochron plot. The age spectra of samples of this type show the inverse staircase pattern and seem to suggest the occurrence of recoil effect. Samples of this type often show the texture where very fine K-rich phases are adjacent to a K-poor phase. Samples of Type 5 have the data distribution with the small dispersion in the inverse isochron plot. This data distribution may be a result of similar degassing of radiogenic and trapped components from similar trapping sites. Further, Ar release patterns of the Deccan Traps samples indicate a peak of non-radiogenic component of Ar at the temperature around 800�Ž. This peak may represent the degassing from some phases in which Ar is dissolved. Candidates for such phases are mesostasis including glass, secondary minerals such as calcite and others. If these nonradiogenic peaks indicate the presence of materials which decompose at this temperature, isotopic data of fractions of less than the decomposition temperature might have been affected by the occurrence of such materials.
Based on the above observations, I have proposed some criteria for sample selection and data treatment for the 40Ar-39Ar dating as follows. (1) It is recommended to use fresh samples and minimize the use of samples which contain alteration materials and mesostasis. (2) Fine-grained holocrystalline rocks are preferable to get reasonable 40Ar-39Ar ages. (3) In order to obtain a more reliable 40Ar-39Ar age, in the age calculation it is better to exclude the isotopic data which are obtained from the lower temperature fractions than a certain temperature around at 800�Ž where the high 36Ar release is observed.
By using the newly developed 40Ar-39Ar dating system and the criteria which have been established as mentioned above, 40Ar-39Ar dating was performed for the Deccan Traps samples. 40Ar-39Ar dating results so far reported are mostly concentrated in the Western Ghats area where thick lava flows crop out and their stratigraphy is well established. In this thesis, I have dated samples from the whole area of the Deccan Traps to reveal the geochronological features of the Deccan volcanism as a total sequence. To monitor the present results by the previous ones, samples from the Western Ghats section were also dated.
Major results are as follows. (1) Flow samples from the Western Ghats area and the central to the eastern sections yield ages ranged from 68 Ma to 64 Ma. Combining with the magnetostratigraphical results, these results indicate that about 70 % of lava flows extruded within the relatively short duration(< 1 m.y.) in the Deccan Traps. (2) 40Ar-39Ar dating results of the Mandla Trap yield older ages(76�`70 Ma) from the Nagpur-Jabalpur section. Further, lava flows from the eastern part of the Narmada river region show slightly older ages(�`70 Ma) than those from the Western Ghats area and the central to the eastern sections. (3) Younger(�`60 Ma) ages have been obtained from the western part of the Narmada river region. (4) As summarized in (1)�`(3), these results indicate that the regional differences of active periods of the volcanism in the different areas seem to exist in the Deccan Traps. (5) 40Ar-39Ar dating results of the dykes which were collected from the two dyke swarms distributed in the West coast area and the Narmada river region indicate no age difference among these dyke swarms. Since these dyke swarms show different orientations systematically(NS in the West coast area and EW in the Narmada river region), this result might indicate that the orientations of these dykes do not reflect the stress field on regional scale but might represent those on local scale during the Deccan Traps volcanism. (6) Laser heating 40Ar-39Ar dating for the biotite samples from the alkali rocks in the Deccan Traps yield similar and/or younger ages relative to those of the main tholeiitic volcanics. Based on this result together with reported 40Ar-39Ar ages for alkali rocks, it is conjectured that main alkalic volcanism might have occurred at least twice during and after the large tholeiitic volcanic active period in the Deccan Traps.