On the embedding of an affine space into a projective space (Q1573686)

Let \(k\) and \(K\) be commutative fields. An embedding of the affine space \(AG(n,k)\) into the projective space \(PG(m,K)\) is an injective mapping \(\psi\) from the point set of \(AG(n,k)\) to the point set of \(PG(m,K)\) which maps collinear points to collinear points and non-collinear points to non-collinear points. The author shows that for \(|k |\geq 4\) and \(n \geq 2\) or \(|k |= 3\) and \(n \geq 3\) every embedding is induced by a semilinear mapping. Let \(\theta: k \to K\) be the associated field homomorphism and let \(r\) denote the degree of the field extension \(K\) over \(k^{\theta}\). If \(n > m\) then \(r \geq 3\) and if also \(r < \infty\) then \(m > {2(n-m) - 1 \over r -2}\). The case \(|k |= 3\) and \(n = 2\) has been settled by \textit{T. G. Ostrom} and \textit{F. A. Sherk} [Can. Math. Bull. 7, 549-559 (1964; Zbl 0125.38401)] and the case \(|k |= 2\) is equivalent to the study of \(2^n\)-arcs and \(2^n\)-caps.