The front group should give two strong and one weak reflection, the rear four strong. But that alone is not enough to determine whether it's a Petzval, it might be a reversed Tessar! So you must try to determine the approximate curvature of the surfaces, which is when it becomes difficult. It gets easier with practice, though.
Hold a light somewhere in the vicinity of your head. I just sit down in a chair under a lamp. It shouldn't be too close to your eye, a little angular separation is a good thing.
Tilt the lens slightly while observing the reflections. Tip it back and forth.
Do all the reflections move the same way, or is one or more moving in the opposite direction?
The outer surface is usually convex (there are exceptions), so that reflection will move the "normal" way. If there' an internal concave surface, that reflection will move the opposite way.
With little care and a little practice (and a loupe to make sure which reflections belong to which surfaces), you can draw a surprisingly good diagram of the lens. Good enough, at least, to place it in the right "design family".