[FOM] looking for help with terminology

[Neil Tennant]

I wonder whether any fom-er can tell me if there is any reasonably
well-established terminology for the following.

I want to isolate formulae of the form

	P(x_1,...,x_n)

where P is a primitive n-place predicate, and of the form

	y = f(x_1,...,x_n)

where f is a primitive n-place function sign, and where in both cases the
x_i are individual variables.

Formulae of the form P(f(x)) won't qualify; nor will formulae of the form
f(y)=g(x) or of the form y=f(g(x))---to choose just three examples of
foils. Yet of course these three formulae count as *atomic*, on the
standard definition of atomic formulae. 

So I'm looking for a term to designate something narrower than the class
of atomic formulae. 

Any suggestions?

Neil Tennant