A new evolutionary algorithm is developed for the optimisation of optical fibre designs. Most traditional fibre designs are circularly symmetric. In recent years, interest in fibres with more complex symmetries and designs has been stimulated by advances in fabrication methods. Here, the fibre design is derived from a set of genes represented by an arbitrary two dimensional array of real values on a polar coordinate grid. The design is obtained from the genes by an averaging process determined by a master gene that controls the symmetry of the final fibre. The interaction of the gene array with the master gene allows designs with different symmetries to lie latent within the same genome. Preliminary analysis of the behaviour of the algorithm is shown for the minimisation of some simple test objective functions.