Research notes on every problem, and a library of the papers behind them. Built from the open erdos repository.
Updated
Burr 1978 ramsey minimal graphs multiple copies
conjecture_p194: The 1978 conjecture giving the size Ramsey number of two arbitrary star forests as a sum of diagonal maxima of star sizes.
corollary_8: A graph arrowing (mK_2, nK_2) with m >= n has a matching of at least [r(mK_2, nK_2)/2] edges; the print omits the integer part, which matters when n is even.
corollary_9: A graph arrowing (mK_3, nK_3) contains at least [r(mK_3, nK_3)/3] vertex-disjoint triangles, and the complete graph on r(mK_3, nK_3) vertices shows the bound is sharp.
theorem_1: The exact size Ramsey number of two uniform star forests and the list of graphs attaining it.
theorem_5: A graph arrowing (mG, H) contains a number of vertex-disjoint copies of G fixed by the orders of G and H and the independence number of H, with Corollaries 6 and 7 as its specializations to (mG, nH).
S. A. Burr, P. Erdős, R. J. Faudree, C. C. Rousseau, R. H. Schelp: Ramsey-minimal graphs for multiple copies, Nederl. Akad. Wetensch. Indag. Math. 40 (1978) no. 2, 187--195 (MR 58 #5387; Zentralblatt 382.05043).
The same paper is Nederl. Akad. Wetensch. Proc. Ser. A 81 (1978), 187-195, DOI 10.1016/S1385-7258(78)80009-2 (the Proceedings numbering, used by Crossref and by Davoodi et al.; the Indagationes numbering is volume 40), communicated December 17, 1977. The copy read for this card is a 9-page OmniPage scan of the article (printed p. n is PDF p. n-186) whose OCR text layer is unreliable for relation signs; the statements below were read on the page images of pp. 187, 188 and 194, the index range of Theorem 1 on 300 and 600 dpi crops. No notice is printed in the scan; the publisher's page could not be read on 2026-10-02 (ScienceDirect returned HTTP 403), and the Crossref record for DOI 10.1016/S1385-7258(78)80009-2 (read 2026-10-02) names Elsevier BV as publisher and lists only Elsevier's text-and-data-mining user license and its open-archive user license (https://www.elsevier.com/open-access/userlicense/1.0/), the publisher's terms and not a reuse grant, every other right reserved.
Read status: claims checked for Theorem 1 with its two arrowing constructions (p. 188), Theorem 5 with Corollaries 6 and 7 (p. 193), Corollaries 8 (p. 193) and 9 (p. 194), and the star-forest conjecture (p. 194); no proof checked.
Writing F -> (G,H) when every red-blue edge-coloring of F yields a red G or a blue H, and R(G,H) for the class of such F, the paper studies the size Ramsey number r-hat(G,H), the minimum number of edges of a graph in R(G,H), introduced by Erdős, Faudree, Rousseau and Schelp. Theorem 1 evaluates it for multiple stars: r-hat(mK_{1,k}, nK_{1,l}) = (m + n - 1)(k + l - 1) for all positive k, l, m, n, and moreover characterizes the extremal graphs, showing that a graph G with G -> (mK_{1,k}, nK_{1,l}) and exactly (m + n - 1)(k + l - 1) edges must be (m + n - 1)K_{1,k+l-1}, or, when k = l = 2, tK_3 ∪ (m + n - t - 1)K_{1,3} for some 1 <= t <= m + n - 1 (both relation signs are the scan's slanted "less than or equal" glyphs, and t = m + n - 1, the union of triangles, is extremal; Davoodi et al. restate the range as 1 <= l <= s + t - 1). The proof supposes a minimal counterexample class C_{k,l} and shows via structural lemmas on its members that it is empty. The second part asks, if F -> (mG, nH), how many disjoint copies of G (or H) must F contain, giving general upper and lower bounds and exact answers in special cases. Problem 561 is this paper's conjecture in its "Questions" section (p. 194): for F_1 = union_{i<=s} K_{1,n_i} with n_1 >= ... >= n_s and F_2 = union_{j<=t} K_{1,m_j} with m_1 >= ... >= m_t, r-hat(F_1, F_2) = sum_{k=2}^{s+t} l_k where l_k = max{n_i + m_j - 1 : i + j = k}; the paper notes that for equal star sizes the conjectured value agrees with Theorem 1.
Source: https://users.renyi.hu/~p_erdos/1978-38.pdf.
Bears on. #561: the conjecture of p. 194 is the problem's statement in its original source, and Theorem 1 proves the problem's formula in the case where all stars of have one size and all stars of have one size. Theorem 5 and Corollaries 6--9 bear on no problem page of this corpus.
Results to transcribe.
- Theorem 1 (p. 188): r-hat(mK_{1,k}, nK_{1,l}) = (m+n-1)(k+l-1); the graphs attaining it are (m+n-1)K_{1,k+l-1}, plus tK_3 ∪ (m+n-t-1)K_{1,3} for 1 <= t <= m+n-1 when k = l = 2.
- Preliminary arrowing bounds (p. 188): (m+n-1)K_{1,k+l-1} -> (mK_{1,k}, nK_{1,l}) and tK_3 ∪ (m+n-t-1)K_{1,3} -> (mK_{1,2}, nK_{1,2}) for 1 <= t <= m+n-1, giving the upper bound on r-hat.
- Conjecture (p. 194): for arbitrary star forests F_1, F_2 as above, r-hat(F_1, F_2) = sum_{k=2}^{s+t} l_k with l_k = max{n_i + m_j - 1 : i + j = k}.
- Theorem 5 (p. 193): if F -> (mG, H) then F contains t disjoint copies of G, t = [(m|V(G)| + |V(H)| - beta_0(H) - 1)/|V(G)|]; Corollaries 6 and 7 (p. 193) specialize it to (mG, nH) and (mG, nG), and the second question of p. 194 asks whether F -> (nG, nG) forces [r(nG, nG)/|V(G)|] copies.
- Corollary 8 (p. 193): F -> (mK_2, nK_2) with m >= n forces a matching of r(mK_2, nK_2)/2 edges as printed, true in the form [r(mK_2, nK_2)/2] and sharp; read literally it fails for even n.
- Corollary 9 (p. 194): F -> (mK_3, nK_3) forces [r(mK_3, nK_3)/3] vertex-disjoint triangles, and this is sharp.
No file of this source is held: no license on record permits its redistribution, and the card cites the edition it names above.