In connection with the thesis of Ch. Charitos (1989), T. Hasanis posed the following question.

The existence of global solutions to the discrete coagulation equations is investigated for a class of coagulation rates of the form ai, j = rirj + αi, j with αi, j≤Krirj. In particular, global solutions are shown to exist when the sequence (ri) increases linearly or superlinearly with respect to i. In this case also, the failure of density conservation (indicating the occurrence of the gelation phenomenon) is studied.

It is shown that an integral domain R has the property that every pure submodule of a finite direct sum of ideals of R is a summand if and only if R is an h-local Prüfer domain; equivalently, (J + K:I) = (J:I) + (K:I) for all ideals I, J and K of R. These results are extended to submodules of the quotient field of an integral domain.

In this paper, it is proved that, for any m unit vectors. x1…, xm in any n-dimensional real Hilbert space, there exists a unit vector x0 such that

for any y∈Sn−1. The exact value of the above integral is calculated, and these results used to improve some lower bounds for multilinear forms on real Hilbert spaces. An integral expression is also given for the complex case.

Consider the equation

where , . The inversion problem for (1) is called regular in Lp if, uniformly in p∈[1, ∞] for any f(x)∈ Lp(R), equation (1) has a unique solution y(x)∈ Lp(R) of the form

with . Here G(x, t) is the Green function corresponding to (1) and c is an absolute constant. For a given s∈[l, ∞], necessary and sufficient conditions are obtained for assertions (2) and (3) to hold simultaneously:

(2) the inversion problem for (1) is regular in Lp;

(3) for any f(x)∈LS(R).

A new and more geometric proof is obtained of Stanley's lower bounds on the face numbers of centrally symmetric simple polytopes.

In [4], we investigated the spaces of continuous functions on countable products of compact Hausdorff spaces. Our main object here is to extend the discussion to arbitrary products of compact Hausdorff spaces. We prove the following theorems in Section 3.

We are interested in the distribution of those zeros of the Riemann zeta-function which lie on the critical line ℜs = ½, and the maxima of the function between successive zeros. Our results are to be independent of any unproved hypothesis. Put

In this paper it is proved that, for x sufficiently large, every integer m with

can be written as m = Σ1≤n≤xεn/n, where εi, = 0 or 1.

This paper treats the reducibility of the quasiperiodic linear differential equations

where A is a constant matrix with multiple eigenvalues, Q(t) is a quasiperiodic matrix with respect to time t, and ε is a small perturbation parameter. Under some non-resonant conditions, rapidly convergent methods prove that, for most sufficiently small ε, the differential equations are reducible to a constant coefficient differential equation by means of a quasiperiodic change of variables with the same frequencies as Q(t).

In 1984 Heath-Brown [2] proved the conjecture of Erdős and Mirsky [1], and obtained the following result.

Theorem. There are infinitely many integers n for which d(n+ 1)= d(n). Indeed, for large x, the number of such n≤x is at least of order x ln-7 x.

Given a Banach space X and a norming subspace Z⊂X*, a geometrical method is introduced to characterize the existence of an equivalent σ(X, Z)-lsc LUR norm on X. A new simple proof of the Theorem of Troyanski: every rotund space with a Kadec norm is LUR renormable, and a generalization of the Moltó, Orihuela and Troyanski characterization of the LUR renormability, are provided without probability arguments. Among other applications, it is shown that a dual Banach space with a w*-Kadec norm admits a dual LUR norm.