It has been shown by Mirsky and Perfect [1] that the theorem of Rado [2], linking matroid theory and transversal theory, has important applications in combinatorial theory. In this note I use it to obtain necessary and sufficient conditions for two families of sets to have a common transversal containing a given set, and then I show how it may be used to obtain a variant of a well-known theorem that was obtained by Hoffman and Kuhn [3] using linear programming methods.

If θ is a real algebraic number of degree r ≥ 2, there is a computable number c = c(θ) > 0 such that

for all rational numbers p/q (q > 0). This follows directly from the definition of an algebraic number, as was shown by Liouville in 1843; and if r = 2 there is no more to be said. Axel Thue was the first to prove a stronger result when r ≥ 3; he showed that if

there are at most finitely many rational numbers p/q that satisfy

Let S be a set of points in n-dimensional space, and suppose that an open sphere of unit radius is centred at each point of S. Suppose that no point of space is an inner point of more than two spheres. We say that S provides a double packing for spheres of unit radius. We define δ2(S), the density of this double packing, to be

where Jn is the volume of a sphere of unit radius, and Nt(S) is the number of points of S inside a cube of side 2t, centred at the origin O. We define δ2, the density of closest double packing, to be

where the supremum is taken over all sets S with the property described above.

As a consequence of the methods developed in the earlier papers of this series [1, 2, 3], an effective algorithm has recently been established for solving many Diophantine equations in two unknowns (see [4,5,6,7]). The algorithm leads to an explicit bound for the size of all the solutions, and, in principle therefore, it enables any specific equation of the type considered to be fully resolved by a finite amount of computation. On examining the various estimates occurring in the course of the exposition, however, it at once became apparent that the computation would involve a very large number of operations and would scarcely be practicable even with a modern machine. It was clear, on the other hand, that a modified version of the fundamental inequality involving the logarithms of algebraic numbers would much facilitate the computational work, and it was in the light of this observation that the researches discussed herein were begun. The object has been to obtain a theorem of an essentially practical nature which may be found useful in application to a wide variety of different problems. The result which we shall establish is neither the most precise nor the most general that can be obtained in this direction, but it would seem to be the most serviceable of its kind, and it would apparently make feasible many calculations which would otherwise have seemed quite out of the question.

The Poincaré problem for the normal modes of oscillations of an inviscid, incompressible fluid contained in an infinitely long cylinder rotating about a direction perpendicular to its axis is investigated.

Let (X, ≺) denote a non-empty (partially) ordered set, or more generally a non-empty set X with an arbitrary transitive relation ≺ on it. The relation ≺ will be fixed throughout what follows, so to simplify the notation we often write (X, ≺) as X. A successor of x ∈ X is an element y ∈ X such that x ≺ y; thus x may or may not be a successor of itself. As usual, a subset A ⊂ X is cofinal if each x ∈ X has a successor in A. A partition of X is a family of (pairwise) disjoint nonempty subsets of X whose union is X.

It was conjectured by Artin [1] that each non-zero integer a unequal to +1, −1 or a perfect square is a primitive root for infinitely many primes p. More precisely, denoting by Na(x) the number of primes p ≤ x for which a is a primitive root, he conjectured that

where c(a) is a positive constant. This conjecture has recently been proved by C. Hooley [2] under the assumption that the Riemann hypothesis holds for fields of the type .

When the motion of a viscous fluid around a gas bubble is discussed, it is frequently assumed, especially for flows at low Reynolds numbers, that the bubble takes on a spherical shape in three dimensions or a circular cross-section in a two-dimensional flow. If this assumption is made, arid the gas within the bubble is assumed to have negligible density and viscosity, then the problem of finding the exterior flow is mathematically overdetermined and it is not obvious that a solution to the problem exists. Moreover, if such a solution does exist, then the over-determination of the system should, in general, give rise to relationships between the flow parameters, that is, certain conditions must be satisfied to ensure the existence of a solution. It is the purpose of this paper to derive these conditions in the case of a two-dimensional Stokes flow. The problem is generalised to the extent that part of the circular boundary is taken to be rigid, on which the no-slip condition is to be satisfied and part is to be a free streamline, on which stress conditions are to be satisfied. The conditions for the existence of a solution to this problem are derived and the solution is found in closed form. The method of solution is that of reducing the problem to one of a mixed boundary-value problem in analytic function theory. The classical solutions for the Stokes flow around a circular bubble and around a rigid circle are then easily derived as limiting cases.

Some diophantine equations in three variables with only finitely many solutions, 113–120. In formula (1) of Theorem 1 and in Corollary 1 read “min” fir “max”

The higher dimensional concepts corresponding to trees are developed and studied. In order to enumerate these 2-dimensional structures called 2-trees, a dissimilarity characteristic theory is investigated. By an appropriate application of certain combinatorial techniques, generating functions are obtained for the number of 2-trees. These are specialized to count those 2-trees embeddable in the plane, thus providing a new approach to the old problem of determining the number of triangulations of a polygon.