Published online by Cambridge University Press: 12 March 2014
The aim of this paper is to correct an error in the proof in [2] of a strengthened version of Hilbert's second ε-theorem.
Hilbert's original theorem says (effectively) that formulae of the form ∃xA → A(εxA) can be eliminated from deductions of ∈-free sentences B (i.e. those which do not contain the ε-symbol) from collections X of ε-free sentences. The extended version (Theorem III. 11 of [2]) says that, in addition, instances of Ackermann's schema can also be eliminated.
The error in [2] occurs in the proof of Theorem III.7 and is described as follows. If A′ is the formula obtained from A by replacing every occurrence of ∃yB by B(εyB) (or ∀yC by C(εyB) if B is of the form ¬ C), and if A is a Q3 or Q4-axiom then, contrary to the claim on p. 73 of [2], A′ is not necessarily an axiom of the same form. For example, if A is the Q3-axiom
where the rank of εx∃yD(x, y) is ≤ the rank of εyB (see p. 70) and where B is D(t,y), then A′ is
which is not a Q3-axiom. In other words, the notion of rank as defined in [2] is inadequate for the proof of Theorem III.11.
The material in this paper is contained in the author's doctoral thesis (Bedford College, London, 1973) which was written under the supervision of Dr. G. T. Kneebone. The author would like to thank the S.R.C. for financial assistance.