The work of Chotiros [J. Acoust. Soc. Am. 97, 199–214 (1995)] has stimulated renewed interest in detecting the Biot type II (slow) P wave for the purpose of ocean bottom characterization. Chotiros postulates that total internal reflection of the type I (fast) P wave can occur at a water–sediment interface and that the observed refracted energy below the critical angle is associated with the type II (slow) P wave. The choice of parameters used in the Biot theory to model these observations is discussed. Two major differences between the parameter set proposed by Chotiros and previous parameter sets for sandy sediments exist. First, the value for the unjacketed bulk modulus, Kr, is about five times less than the bulk modulus for quartz crystals, which is commonly used. This value is obtained by equating it to a value of effective grain modulus measured by Molis and Chotiros [J. Acoust. Soc. Am. 91, 2483(A) (1992)]. Second, the value for the frame bulk modulus, Kb, obtained by fitting to measured values of the type I (fast) P wave speed, is from 50 to 100 times larger than in the previous models. The larger value for the frame bulk modulus is in part a consequence of the smaller value of the unjacketed bulk modulus. Within the constraints imposed by the Biot theory, subject to the assumption of an effective homogeneous solid, the proposed values of unjacketed bulk modulus and frame bulk modulus are incompatible. It is shown here that it is not possible to predict Chotiros’ observed large phase speed of the type II P wave by assuming an equivalent homogeneous solid. However, it is possible to choose a compatible set of parameters for the Biot theory so as to fit the velocity observations once the assumption of an equivalent homogeneous solid is relaxed.