We are starting to follow up the work of John Shrimpton' Bangor thesis, whose aim was the application of categorical methods in group theory, and was partially published as "Some groups related to the symmetry of a directed graph" JPAA 72 (1991) 303-318, and would be grateful for information on publications in this area. We know of the paper by Bumby and Latch, Internat J Math Sci 9 (1986) 1-16, P Ribenboim Algebraic structures on graphs, Alg. Univ. 16 (1983) 105-123; Bill Lawvere, Qualitative distictions between some toposes of generalized graphs, Cont. Math. 92, 261-299, (1989) and would be grateful for further information. The aim of the project is to use categorical methods to give insight into combinatorial problems, constructions and questions. A further aim was to understand Grothendieck's notion of "Teichmuller groupoid" referred to in "Pursuing stacks", in the sense that if what he asserted could be done was actually written up for surfaces, one should also be able to do it for graphs, (or vice versa), and so obtain presentations of a "symmetry groupoid" of a graph from some decomposition into smaller pieces. I have no idea how to do this. Grothendieck's observation in this area is quoted in my survey "From groups to groupoids" Bull London Math Soc. 19 (1987) 113-134. Part of this programme is "what should be the symmetry object of a given structure?". New answers should be of general interest, for obvious reasons. Shrimpton's work showed that these questions lead to that of, for example, inner automorphism of a directed graph, and the interpretation of this in graph theoretic terms. There should be more in this area. Ronnie Brown Prof R. Brown School of Mathematics, University of Wales, Bangor Gwynedd LL57 1UT UK