SIR,-We would like to present the results of a reversed seismic refraction line shot between Thorns Beach (SZ39309610), on the mainland of Hampshire and Elmsworth Brickworks (SZ46209228),nearNewtown on the Isle of Wight, a distance of about 5 km.Reversal was obtained by firing, in either direction, a series of shots across the Solent, the geophone spread being laid on land. The technique adopted was similar to that described by Griffiths et al, (1961).A total of twenty shots was fired at sea, eleven on the line shot towards Thorns Beach and nine on the line shot towards Elmsworth Brickworks. Charges varied in weight from 0.5 kg to 2 kg. Water depths beneath each shot were recorded. The shot positions were fixed by theodolite bearings observed from two survey stations situated on land at each end of an accurately measured base line. Error in readinj the bearings is estimated to be of the order of 2 minutes of arc. Thus at a distance of 3km the error in shot position is of the order of 6.5 m. This is considered to be one of the principal sources of error in this survey.The time-distance graph obtained for the reversed line is shown in Figure 1. To simplify presentation the arrival data plotted is for the geophone nearest to the shot. A correction has been applied to these arrival times for variation of water depth.No direct arrivals were observed, but the time for the shot nearest to either end of the line can be used to give a limiting maximum velocity for the top layer. In both cases this is calculated to be less than 1,700 m/sec (5,600 ft/sec). A short reversed line fired at Thorns Beach gave a near surface velocity of 1,660'i 6 m/sec (5,450i20 ft/sec). This velocity is used for the surface layer (layer 1) in the interpretation.The apparent velocities and their intercepts together with standard errors are summarized in Table 1. Figure 2 presents our interpretation of the depths to the refractors. This indicates that the three main refractors rise slightly towards the mainland.The lowermost refractor (layer 3/layer 4 interface) we consider to be the Tertiary-Cretaceous unconformity, which Wooldridge & Linton (1955) have estimated to be about 500 m deep in this area.Layejr 4 with velocity of 3,500 m/sec must consequently be the Chalk. Laboratory determinations by Laughton & Stride (1957), for a stratigraphic profile within the Chalk at Flamborough Head, indicate a velocity range from 3,000 m/sec to 5,000 m/sec. More recently, commercial seismic investigations in the North Sea (Cook, 1965) reveal a congressional wave velocity for the Chalk of approximately 3,650 m/sec. These velocities are appreciably higher than velocities for the Chalk obtained by Bullard et al. (1940) in eastern England (1,930-2,520 m/sec), and Day et al. (1956) in the English Channel and the Western Approaches (2,290-2,440 m/sec). Stride (1959) suggested that a velocity of 3,000 m/sec observed in refraction surveys over the Dogger Bank may represent the Chalk, but could include Upper Palaeozoic and/or other Mesozoic rocks.The velocities for th...