  The two prior distributions can have either common or different precisions (variances).  In either case, these precisions (variances) can be known or not.
  Each prior distribution (ENUM) must be specified through a multi-dimensional list as described below. Some dimensions are optional, depending on the context.
  \tabular{ll}{    \code{mu0} \tab The prior mean for each group.
      Numeric vector of length 2.\cr
    \code{n0} \tab The prior sample size equivalence from each group.
      Numeric vector of length 2.\cr
    \code{prec.shape} \tab Gamma shape parameter for the unknown precision within each group.
      Numeric vector of length 2.\cr
    \code{prec.rate} \tab Gamma rate parameter for unknown precision within each group.
      Numeric vector of length 2.\cr
    \code{common.sd} \tab Whether the (unknown) precisions are assumed to be the same in each groups (lambda_1 = lambda_2).
      Default is \code{False}.
      If true, then both \code{prec.shape} and \code{prec.rate} should be of length 1.
      It is not necessary to define \code{common.sd} when both precisions are known (see \code{known.sd} below).\cr
    \code{known.sd} \tab  Values for the standard deviation when the variance/precision is known.
      Optional.
      Numeric vector of length 1 or 2; can be of length 1 if only lambda_1 is known; \bold{must} be of length 2 if lambda_2 is known, in which case \code{known.sd[1]} needs to be set to \code{NA} if lambda_1 is unknown.
      If both precisions (lambda_1 and lambda_2) are known, then \code{prec.shape} and \code{prec.rate} are irrelevant (need not be defined).
      If only second group precision (lambda_2) is known, \code{prec.shape} and \code{prec.rate} can be of length 1 only.
      If only first group precision (lambda_1) is known, \code{prec.shape} and \code{prec.rate} need to be of length 2, but the first component in each of them will be ignored (can be set to \code{NA}s).
  }