What is the origin of the terms fences, funnels, and separatrix in differential equations?
I am not asking what they mean but rather who introduced them and how. They seem to be very recent terms, not used in the classical literature; at the same time, they seem to be taken for granted in some contemporary sources.
I found that Hubbard's paper uses fence and funnel, but it's not clear from reading if he is introducing them or simply referring to them.
I'm pretty sure "separatrix" is older than "fence" and "funnel". For example, Lefschetz "Lectures on Differential Equations" from 1946 contains "separatrix".
[EDIT] It seems "fence" and "funnel" were introduced in 1991 in Hubbard and West's book "Differential Equations: A Dynamical Systems Approach". The Preface says "In order to accomplish this goal, we must introduce some new terminology right at the beginning, in Chapter 1. The descriptive terms "fence," "funnel," and "antifunnel" serve to label simple phenomena that have exceedingly useful properties not exploited in traditional treatments of differential equations."