We present a time series of 8 - 13 $\mu$m spectra and photometry for SN 2014J obtained 57, 81, 108, and 137 d after the explosion using CanariCam on the Gran Telescopio Canarias. This is the first mid-IR time series ever obtained for a Type Ia supernova. These observations can be understood within the framework of the delayed detonation model and the production of $\sim$0.6 $\rm M_\odot$ of $^{56}$Ni, consistent with the observed brightness, the brightness decline relation, and the $\gamma$-ray fluxes. The [Co III] line at 11.888 $\mu$m is particularly useful for evaluating the time evolution of the photosphere and measuring the amount of $^{56}$Ni and thus the mass of the ejecta. Late-time line profiles of SN 2014J are rather symmetric and not shifted in the rest frame. We see Argon emission, which provides a unique probe of mixing in the transition layer between incomplete burning and nuclear statistical equilibrium. We may see [Fe III] and [Ni IV] emission, both of which are observed to be substantially stronger than indicated by our models. If the latter identification is correct, then we are likely observing stable Ni, which might imply central mixing. In addition, electron capture, also required for stable Ni, requires densities larger than $\sim$$1 \times 10^9$ g cm$^{-3}$, which are expected to be present only in white dwarfs close to the Chandrasekhar limit. This study demonstrates that mid-IR studies of Type Ia supernovae are feasible from the ground and provide unique information, but it also indicates the need for better atomic data.