Abstract It is shown that under favorable conditions many quite good reflections can be obtained from within the crust. They are strong enough and numerous enough to suggest that they may originate from several large sills. Details of the evidence and arguments for this necessarily indirect conclusion are discussed. The reflection method is considered to be the only method of obtaining detailed information on the structure and composition of the deeper crust; refraction methods cannot yield adequate detail. The angle of reflection should be kept as small as possible to stay near the reflection maximum at normal incidence, and the receivers should be offset far enough that the decay of surface wave noise from the source permits reflections to be seen. Appendixes treat strike and dip computation, normal time corrections used for correlations, and waves passing through focuses.