strange results from some predicate-arity calls

[AdrickTench]

I'm not sure if this is a bug in the code for predicate-arity or a result of some strange definitions somewhere.

Calling predicate-arity on some built-ins yields strange results, for example:

metta+>!(predicate-arity eval)
[2, 1, 6, 4, 3, 2, 6, 4]
metta+>!(predicate-arity chain)
[4, 2]

I would expect to see

metta+>!(predicate-arity eval)
[2]
metta+>!(predicate-arity chain)
[4]

[TeamSPoon]

Very important to read about the changes of expected behaviour in #177