arXiv++ Combinatorics

This paper characterises the structure of every maximal weak or strong Gallai-Schur partition. The results confirm the exact values of Gallai-Schur numbers provided by Budden (2020) in the strong case, and provide corresponding values for weak Gallai-Schur numbers. The proofs are elementary and standalone.

This paper sets out the results of a range of searches for linear and cyclic graph colourings with specific Ramsey properties. The new graphs comprise mainly 'template graphs' which can be used in a construction described by the current author in 2021 to build linear or cyclic compound graphs with inherited Ramsey properties. These graphs result in improved lower bounds for a wide range of multicolour Ramsey numbers. Searches were carried out using relatively simple programs (written in the language `C') to generate clauses for input to the PeneLoPe and Plingeling parallel SAT-solvers. When solutions were found, the output from the solvers specified the desired graph colourings. The majority of the graphs produced by this work are `template graphs' with parameters in the form $(k,k,3)$ or $(k,l,3)$ with $k \ne l$. Using these template graphs in familiar constructions, it has been possible to demonstrate significant improvements for lower bounds for most $R_r(k)$ for $5 \le k \le 9$ and $r \ge 4$. These improvements provide correspondingly increased lower bounds on $Γ(k) = \lim_{\substack{r \rightarrow \infty}} R{_r}(k)^{1/r}$. We also show that $R_3(8) \ge 7174$ and $R_3(9) \ge 15041$. Other new lower bounds include $R(3,6,6) \ge 338$ and $R(3,8,8) \ge 941$, based on non-template cyclic graphs, and the interesting particular cases $R(3,4,5,5) \ge 729$ and $R(3,5,5,5) \ge 1429$. A spreadsheet containing specimens of many of the graphs mentioned here will be attached as an ArXiv ancillary file.

In this simple paper, we exhibit a Schur partition giving rise to a triangle-free linear colouring of $K_{1697}$ in 7 colours. Thus we show that the Schur number $S(7) \ge 1696$ and the multicolour Ramsey number $R_{7}(3) \ge 1698$. We also demonstrate a specific partition of the closed integer interval $[1, 376]$ which, through repeated use of a specific construction, allows us to conclude that $S(k+5) \ge 376.S(k)+160$ for any positive integer $k$. The existence of this partition, with its specific properties, implies that $\lim_{\substack{r \rightarrow \infty}} R{_r}(3)^{1/r} > 3.273$.

This paper records some apparently new results for the partition of integer intervals [1, n] into weakly sum-free subsets. These were produced using a method closely related to that used by Schur in 1917. New lower bounds can be produced in this way for partitions of unlimited size. The asymptotic growth rate of the lower bounds, as the number of subsets increases, cannot be less than the same growth rate for strongly sum-free partitions, and therefore exceeds 3.27. Specific results for partitions into a 'small' number of subsets include $WS(6) \ge 642$, $WS(7) \ge 2146$, $WS(8) \ge 6976$, $WS(9) \ge 21848$, and $WS(10) \ge 70778$.

In 1952, J.H.Braun claimed to have established a formula giving a lower bound for certain partitions of sets of integers into weakly sum-free classes. However, no proof or supporting construction was published at that time. In today's terminology, that claim was equivalent to giving a formulaic lower bound for the weak Schur number $WS(s)$. $WS(s)$ is the maximum number such that there exists a weak Schur partition of the integers from 1 to $WS(s)$, into $s$ subsets. In a weak Schur partition of a set of integers, there can be no three distinct members $a$, $b$ and $c$ in any subset, such that $a+b=c$. An iterative construction described in this paper results in a similar formulaic lower bound. Although different from that given by Braun, it reproduces the result $WS(6) \ge 554$ implied by his formula, and exceeds it for all larger values of $s$. Various starting points can be used as a basis for the iterations. This result itself is no longer remarkable: it has been proven elsewhere that $WS(6) \ge 642$. Even so, it is hoped that the formula and its underlying construction may nevertheless be of interest to those interested in weak Schur partitions and/or the closely-related linear Ramsey graphs.

A construction described by the current author (2017) uses two linear prototypes to build a compound graph with Ramsey properties inherited from the prototype graphs. The resulting graph is linear; and cyclic if both prototypes are cyclic. However, it will not generate a cyclic graph from a general linear prototype. Building on the properties of that construction, this paper proves that a general linear prototype graph of order m can be extended using a single new colour to produce a new cyclic graph of order $3m - 1$ which is triangle-free in the new colour, and has the same clique-number as the prototype in every other colour. The paper then describes a cyclic Ramsey $(3;3;4;4; 173)$-graph derived by constrained tree search, thus proving that $R(3;3;4;4) \ge 174$. Using a quadrupling construction to produce a further cyclic graph, it is shown that $R(3;4;5;5) \ge 693$. A compound cyclic Ramsey $(3;7;7; 622)$-graph derived by a limited manual search is then described. Further construction steps produce a $(8;8;8; 6131)$-graph, showing that $R_3(8) \ge 6132$. The paper concludes by showing that $R_4(7) \ge 81206$ and $R_4(9) \ge 630566$, implying corresponding improvements in the lower bounds for $R_5(7)$ and $R_5(9)$ and beyond. These results follow from the existence of cyclic prototype graphs derived by Mathon-Shearer 'doubling'.

A construction described by the current author in 2017 uses two linear `prototype' graphs to build a compound graph with Ramsey properties inherited from the prototypes. This paper describes a generalisation of that construction which has produced improved lower bounds in many cases for multicolour Ramsey numbers. The resulting graphs are linear, as before, and under certain specific conditions set out here, they can be cyclic. The mechanism of the new construction requires that the first prototype contains a triangle-free `template' in one colour, with defined properties. This paper shows that in the compound graph, clique numbers in the colours of the first prototype may exceed those of the prototype. However, it proves necessary only to test for a limited subset of the possible cliques using these colours, in order to evaluate the relevant clique numbers for the entire graph. Clique numbers in the colours of the second prototype are equal to those of that prototype. These attributes enable the efficient searching of a new family of graphs based on the `template' approach. It has also been found that there are a number of useful cases in which the clique numbers of the first prototype are not increased by the compounding process. As a result of this construction many lower bounds can be improved. The improvements include $R_4(5) \ge 4073$, $R_5(5) \ge 38914$, $R_3(6) \ge 1106$, $R_4(6) \ge 21302$, $R_4(7) \ge 84623$ and $R_3(9) \ge 14034$. It is also shown that $R_3(9) \ge 14081$ using a non-linear construction.