Seismic arrays have advantages over a worldwide network of single instrument stations because (1) They provide direct measurements of dT/dA over a limited range of and enable resolution of minor perturbations in the travel time curve; (2) Enhancement of signalto noise ratio (SNR), allows the detection of weak signals, which might be missed on single instruments and (3) The later arrivals are detected and analysed on the basis of their apparent velocities to include them in the inversion of the slowness (inverse of apparent velocity) and travel time data. In array measurements of dT/d it is as necessary to allow for the effects of crust and upper mantle structure as it is in travel time studies. However, because of the reduced dimensions of the array compared to a large network of stations, the accuracy with which these effects are required to be known is much greater. Also, seismic arrays are prone to systematic errors in the determination of slowness and, if care is not taken, array data become somewhat less reliable for the determination of velocities than are travel time data, where some averaging of station anomalies may be expected to occur. Whereas travel time studies provide an average velocity structure over a large area of observations, array measurements allow derivation of velocity models which are better representatives of particular regions. Array measurements have led to the determination of the deviations of the velocity structure from the classical velocity distributions of Jeffreys and Gutenberg and confirmed some results of surface wave dispersion studies, which had demanded, for example, the existence of an S low-velocity layer, at least in some regions (Press, 1959; Takeuchi et al., 1959; Dorman et al., 1960) and two zones of rapid velocity increase at depths near 350-450 km and 600-700 km (Anderson and Toksoz, 1963 and Toksoz and Anderson, 1966; Toksoz et al., 1967). Also, surface wavev dispersion results differ for paths through the oceans, shields and tectonic regions so that the concept of a spherically symmetric earth is no longer tenable (Kanamori, 1970; Dzeiwonski, 1971). However the fine structure (e.g. the number of velocity discontinuities, their depths and velocity gradients between various layers) has varied between various studies (see, for instance, Johnson, 1967; Green and Hales, 1968; Kaila et al., 1968; Helmberger and Wiggins, 1971; Masses et al., 1972 Simpson et al, 1974; Ram Datt and Muirhead, 1976, 1977; Ram and Mereu, 1977; Ram Datt, 1977; Hales et al., 1978). The causes of the observed variations include (i) regional variations (ii) lack of data in certain epicentral distance ranges and (iii) differences in the procedures of analysis and interpretation of data. For the lower mantle, though it is believed to be more homogeneous, deviation from homogeneity were reported by various authors on the basis of first arrival data (Chinnery and Toksoz, 1967; Wright, 1968; Archambeau et al., 1969; Johnson, 1969; Corbishley, 1970; Wright and Cleary, 1972).