Stoichiometric defects (Schottky and Frenkel defects) in ionic crystals are formed due to irregularities in the ionic arrangement which maintain the stoichiometry and electroneutrality of the solid. It is known that $\ce{AgBr}$ shows both Frenkel and Schottky defects while most other compounds show only one of these defects; for example, $\ce{AgCl}$ shows only Frenkel defects.
The ratio of the sizes of the cation and anion determines which type of defect is shown by the crystal. If the ratio is closer to unity, Schottky defects are more common, and if it deviates significantly from one, Frenkel defects are more common. Intuitively, this is correct since Frenkel defects involve the movement of the cation (or the smaller ion) into interstitial sites, which is indeed easier if the ion is very small as compared to the other ion, making it more mobile. Generally, the activation energy for cationic movement is lower. And if the ionic sizes are comparable, filling the interstitial sites is not feasible, and hence Schottky defects are more likely to be exhibited by the crystal.
The $\ce{Cl-}$ ion is smaller than $\ce{Br-}$ (i.e. the ratio $r_+/r_-$ for $\ce{AgCl}$ is closer to 1 than it is for $\ce{AgBr}$). Yet, $\ce{AgBr}$ displays Schottky defects while $\ce {AgCl}$ does not. Why is that so? (The size and hence the mobility of $\ce{Ag+}$ should be the same in both cases.)