A set D of k-letter words is called a comma-free dictionary (2), if whenever (a1a2 . . . ak) and (b1b2 . . . bk) are in D, the "overlaps" (a2a3 . . . akb1), (a3a4 . . . akb1b2), . . . , (akb1 . . . bk-1) are not in D. We say that two k-letter words are in the same equivalence class if one is a cyclic permutation of the other. An equivalence class is called complete if it contains k distinct members. Comma-freedom is violated if we choose words from incomplete equivalence classes, or if more than one word is chosen from the same complete class.

Let be the class of m × n matrices all of whose entries are either 0 or 1 where every matrix A in the class satisfies the conditions that row i of A has ri ones, i = 1, 2, . . . , m; and column j of A has sj ones, j = 1, 2, . . . , n. We let R = (r1 . . . , rm), S = (s1, . . . , sn), and assume that r1 ≥ r2 ≥ . . . ≥ rm ≥ 0; s1 ≥ s2 ≥ . . . ≥ sn > 0. When this is the case we say the class is normalized.

We shall consider the classification problem for space forms of (Riemannian manifolds which are covered by) real, complex, and quaternionic Grassmann manifolds. In the particular case of the real Grassmann manifold of oriented 1-dimensional subspaces of a real Euclidean space, this is the classical "spherical space form problem" of Clifford and Klein. We shall not consider space forms of the Cayley projective plane because it is easy to see that there are no non-trivial ones.

The present study was motivated by an investigation of algebraic conjugates in the complex plane (cf. 4 for one of the results) where some of its concepts are extended and applied.

Let F be a flat (linear subspace) in real affine n-space. The points z = (ζ1, . . . , ζn) on F for which the least common denominator of the coordinates ζv is minimum form a grid G, the main grid of F, studied in § 3.

The purpose of this note is to discuss some properties of singular fibrations introduced by Montgomery and Samelson (4) and studied by Hu (2) and Conner and Dyer (1). This work was motivated by the last paper, where singular fibrations in which the fibre is a sphere were studied. Here more general fibres will be considered.

Let Un be an n-dimensional unitary space with inner product (u, v). For vectors u1, . . . , ur ∊ Un, r ≤ n, let denote the Grassmann exterior product (4) of the ut; it is a vector in Um where m = nCr. If also , then is the determinant of the r X r matrix If A is a linear transformation of Un to itself, the rth compound of A is defined by

If R is a sequentially continuous linear lattice, a function f(x), denned on R+ = {x:0 ≤ x ∊ R} with 0 ≤ f(x) ≤ + ∞, will be called a monotone function if it satisfies

In a previous paper (2), one of us has derived a formula for the counting series for bicoloured graphs.2 These are graphs each of whose points has been coloured with exactly one of two colours in such a way that every two adjacent points have different colours.

In this paper we first enumerate bicoloured graphs without isolated points and connected bicoloured graphs. This leads us to corresponding problems for bicolourable graphs. Such a graph has the property that its points can be coloured with two colours so as to obtain a bicoloured graph.

In the series of "Census" papers, of which this is the fourth, we attempt to lay the groundwork of an enumerative theory of planar maps (12, 13, 14). The maps concerned are rooted in the sense that some edge is fixed as the root, and a positive sense of description and right and left sides are specified for it. This device simplifies the theory by ruling out the possibility of a map being symmetrical.

In 1927, J. H. Redfield (12), discussed some of the links between combinatorial analysis and permutation groups, including such topics as group transitivity, the enumeration of certain geometrical configurations, and the construction of various permutation isomorphs of a given group. Except for a revision of Redfield's treatment of transitivity by D. E. Littlewood (6), this 1927 paper appears to have been overlooked. However, it has recently been described (5) as a remarkable pioneering paper which appears to contain or anticipate virtually all of the enumeration results for graphs which have been discovered and developed during the last thirty years.

Let A be a finite-dimensional associative algebra over a field F. Let R denote the radical of A. Assume that A/R is separable. Then it is well known (the Wedderburn principal theorem) that A possesses a Wedderburn decomposition A = S + R (semi-direct), where S is a separable subalgebra isomorphic with A/R. We call S a Wedderburn factor of A.

Let denote the set of points x = (x1, . . . , xn) with integral co-ordinates in Euclidean n-space. For any fixed integer m ≥ 2, let C = C(m) be the set of such points in the cube 0 ≤ xi < m (i = 1, 2, . . . , n) and let be any subset of C. Suppose that f(x) is any single-valued, integral-valued function, defined for all x ∊ . We consider solutions x ∊ of the congruence

This paper is concerned with the recent work of A. W. Goldie (1 ) on principal right ideal rings (p.r.i. rings). We shall prove some of his main structure theorems using the methods of (3) and (4), and in so doing shall weaken some of his hypotheses.

Let M be a vector space of dimension n over a field K of characteristic ≠2 and f a non-degenerate quadratic form on M. The automorphisms of the orthogonal group On(K,f), the rotation group On+(K,f), and their corresponding projective groups POn(K,f), POn+(K,f), were determined by J. Dieudonné for n sufficiently large under the condition that the index of f is greater than zero (see 2).

Let Fn(k) denote the total number of k-coloured graphs on n labelled nodes and let Mn(k) denote the number of graphs on n nodes that are coloured in at most kcolours ; also let fn(k) denote the number of connected k coloured graphs on n nodes. Read (3) has proved the following formulas:

A matrix is said to be positive definite if it is hermitian and if all of its characteristic values are positive. It is well known, and easy to prove, that the necessary and sufficient condition for a matrix P to be positive definite is that its hermitian quadratic form

with any vector v ≠ 0 be positive. (This will imply, in the present article, that it is real.) It is easy to see from (1) that if P1 and P2 are positive definite, the same holds of a1P1 + a2P2 if a1 and a2 are positive numbers.

In his fine and stimulating book Introduction to geometry (1), Coxeter considers three mutually tangent circles C1 , C2, and C3 (Fig. 1) and calculates the radii of the two circles E1 and E2 that touch all these three. If the centres of Ci are the vertices of a triangle ABC their radii are s — a, s — b, and s — c; α, β, and γ are defined by

This paper is a continuation of the earlier papers (1, 5) in which the author studied matrices with entries from the algebra C() of all continuous, complex-valued functions on an extremely disconnected, compact Hausdorff space . (Such spaces are sometimes called Stonian, after M. H. Stone, who first considered them in (8). They arise naturally as maximal ideal spaces of abelian W*-algebras.) In this note, three theorems are proved.

In a recent paper, Olver (2) obtains a set of formulae that completely determine the asymptotic behaviour of the Hermite polynomials, Hn(z), as n —> ∞ and z is unrestricted. His proof depends on a technique that he has developed for discussing the asymptotics of solutions of second-order, linear, homogeneous differential equations satisfying certain conditions. We believe it fair to say that Olver's work follows the tradition of most of the major theorems of classical asymptotics. The results contained in theorems such as Watson's lemma and Perron's proof of the Method of Laplace are based on an acceptance, on an a priori basis, of the Poincaré type expansion.

In the Mathematical Games department of the Scientific American for November 1958 (12), William H. Tutte gave an entertaining account of the researches of himself and three colleagues (C. A. B. Smith, A. H. Stone, and R. L. Brooks) concerning "squaring the square." The problem is to subdivide a square into a finite number of non-overlapping smaller squares of which no two are the same size. At first suggested as impossible, a number of different solutions have been published. The first one, by Sprague in 1939 (9), showed a square divided into 55 smaller unlike squares.