In an earlier study of turbulent premixed Bunsen flames, we employed a two-component laser Doppler velocimeter (LDV) system and a three-element electrostatic probe to investigate influences of flamelets on local gas velocities in the reaction-sheet regime. Because the two-component LDV does not provide complete information about the gas velocity and the three-element probe does not provide complete information on the direction of flamelet motion, it was not possible to fully explain the results of the LDV measurements correctly. To remedy this difficulty so that the explanations can be completed, in the present work the two-component LDV is replaced by a three-component LDV and the three-element probe is replaced by a four-element probe, which provide full information on the direction of flamelet motion. The new LDV system was employed to measure the three components of velocity at a midflame point on the centerline of a propaneair flame of equivalence ratio 1.10 on a 26-mm-diameter Bunsen burner with fully developed turbulent pipe flow upstream at a Reynolds number of 6700. The LDV measurement volume was 1mm below the leading electrode of the four-element electrostatic probe that provided the velocity and direction of motion of flamelets in the turbulent flame brush. It was found that when the coordinate system was rotated about a vertical axis to the plane containing the flamelet velocity vector, most of the burnt-gas motion following flamelet passage often remained largely in this plane, although in many cases a nearly constant rate of rotation about this plane was detected. The sequence of LDV data for gas motion in the burnt gas consistently showed movement beginning first in a direction different from that of flamelet passage, then gradually rotating until finally becoming the direction of flamelet passage. This behavior is explained qualitatively by considerations of momentum conservation and of the nature of the LDV measurements. Earlier speculations concerning the role of buoyancy are thereby resolved, now showing that buoyancy is not the principle controlling phenomenon, and additional information about turbulence structures in these flames is obtained.