We already have the general solutions to (2),
There are two reasons why (2) is not a general solution to the overall set of equations of motion. First, it has too many arbitrary parameters. Each point on either side of the interface (or molecule, if you prefer to count discrete objects) obeys a second order (in time) equation of motion. The general solution thus must have two arbitrary parameters for each point in space. However, there are four arbitrary functions in (4), each of which can take any value for each value of its argument. Thus, there are four adjustable parameters per point in space, rather than just two. We must therefore be able to eliminate two of the four functions in (4) using some other constraints in order to leave ourselves with just two adjustable functions.
The second difficulty with (4) gives us the additional
constraints to resolve the first difficulty. Although
(4) solves the equations of motion for the interior
points, it does not necessarily solve the equations of motion for the
membrane (1,3). These conditions translate into
relations among the 's and
's which will allow us to
eliminate the extra freedom while satisfying the equation of motion for
all points in the system.