Let us draw a sphere of radius   in such a way that the primary beam passes along diameter IO, and place the origin of the reciprocal lattice at O (this is the Ewald sphere).

Let us rotate the figure on the left: when one reciprocal lattice point lies on the sphere the Bragg's law is verified .

Indeed if the vector r*_H ends on the sphere, the direct lattice planes (perpendicular to r*_H) will lie parallel to IP and will make an angle with the primary beam. Then

AP is the direction of the diffracted waves. (It makes an angle 2 with the primary beam. Therefore we can imagine the crystal located at A).

Therefore, the necessary and sufficient condition for Bragg's equation to be verified for the family of planes (hkl), is that the lattice point defined by the vector r*_H lies on the surface of the Ewald sphere. Any diffraction experiment will then aim at causing contact of the reciprocal lattice points with the Ewald sphere.

If r*_H is larger than the diameter of the Ewald sphere ( r*_H is greater than  2/, or equivalently d_H is less than /2 ) we will be not able to observe the reflection H.

This condition defines the so called limiting sphere , with centre at O and radius 2/: only the lattice points inside the limiting sphere will be able to diffract. The wavelength defines the amount of information available from an experiment.

In ideal conditions, the wavelength should be short enough to leave out of the limiting sphere only those lattice points with intensities close to zero due to the decrease of atomic scattering factors.