If x isa variable in a probabilistic model that takes on values in a set Ω, the prior distribution over x is simply P(x), or the (unconditional) probability thatit takes on any value in Ω.
See Bayesian learning, Likelihood function, Poisson distribution, Posterior distribution