In the Pacific Journal of
Mathematics, 71 (1977), Richman and Walker gave a natural definition for Ext in an
arbitrary pre-abelian category. Their Theorem 4, which states that (αE)β = α(Eβ)
for an arbitrary sequence E, is in error. We show, however, that (αE)β = α(Eβ)
does hold for a stable exact sequence. Without Theorem 4, the crucial step
in their theory is showing that αE is stable if E is stable. We prove this.
Consequently, the theory of Richman and Walker for Ext in a pre-abelian category is
valid.