Some evidence against Fourier analysis as a function of the receptive fields in cat's striate cortex