The problem of neutron diffraction by crystals is treated by analogy with X-ray diffraction, consideration being given to the perfect crystal, the mosaic crystal and the powdered block. The first part of the paper deals with a comparison between X-ray and neutron diffraction and it is shown that quantitatively the two are similar, apart from the case of the thick mosaic crystal where the very low values which are usually found for the true absorption of neutrons result in the integrated reflexion being largely independent of structure factor under practical conditions. The second part deals with the practical problem of diffracting the collimated beam of thermal neutrons from an atomic pile. Again, perfect and mosaic crystals are considered and it is also shown that the powder method, which it is desirable to use when structure factors are to be determined, will be feasible if the number of pile neutrons which hit the mono-chromatizing crystal is greater than 10 5 per sec.