It is pointed out that vertical wave number spectra of gravity waves can be divided into subranges of weak and strong wave interactions and that observed −2.5 to −3 power law spectra may fall within the strong interaction subrange. An elementary criterion is derived to help distinguish between weak and strong waves. It is then shown that the strong wave subrange is predicted by a well-known theory of Lumley. The relation of theory and observation to breaking gravity waves, turbulence dissipation, and length scales is discussed. The validity of Lumley's theory is summarized, and unresolved questions for future investigations of atmospheric wave spectra are raised.