The terms superfluid and normal fluid may be interpreted in various ways, and in Landau's theory the fraction of normal fluid given by flow experiments (e.g., second sound) is not a direct measure of the number of atoms involved in the excitations. Furthermore, the anomaly in the specific heat above the $\ensuremath{\lambda}$ point indicates the continued existence of unexcited material above the $\ensuremath{\lambda}$ point, probably in the form of (nonspherical) droplets or clusters, which does not actually contribute to the superfluidity. These rather complex relationships are considered in the introduction, and it is concluded that, in spite of the various possibilities of choosing the two components for the two-fluid theory, any pair can be considered as thermodynamic components. It is also concluded that it is probable that the normal fluid and superfluid are separated in ordinary space as well as momentum space. There follows a discussion of the equation for second sound, and a comparison of the values for $x$ (mole fraction of normal fluid) obtained from second sound and from the Andronikashvili experiment. The values of $x$ are also discussed in connection with the values of the roton part of the specific heat, ${c}_{r}$. It is shown that these are difficult to reconcile from a thermodynamic point of view on the basis of any of the usual theories, and it may be necessary to reinterpret the equation for second sound in the region where both phonon and roton excitations are of importance. At higher temperatures the apparent anomalies, and especially the rapid rise of $\frac{{c}_{r}}{x}$ near the $\ensuremath{\lambda}$ point, are readily explained with the aid of the unexcited droplets mentioned above. Finally a critical analysis is made of the assumption, inherent in the second-sound equation, that entropy is carried only by normal fluid and not by the superfluid, which is in apparent contradiction with the fact that the entropy of mixing of superfluid and normal fluid cannot be zero if they are separated in ordinary space. It is shown from two points of view that there is no actual contradiction. In the first procedure a pressure is introduced which arises from the forces tending to separate normal and superfluid, and the accompanying work is considered. It is shown that this pressure must be considered to reside in the superfluid. The second procedure starts directly with the energy equation. It is shown that H. London's equation for the fountain pressure follows directly.