Prisican on the Stress of calefacit.

On pg. 127 fn. 16 I mention that calēfacit was probably accented as if two words, i.e. cálē fácit, which allowed the ē to undergo Iambic Shortening.  In support of this I note the failure of the a of facit to weaken and the occasional inversion and/or separation of the two parts. In addition I should have mentioned the explicit testimony of Priscian (Keil 2:402):

si uero facio uerbo uel fio integris manentibus aliud uerbum infinitum ante ea componatur, non solum significationes et coniugationes integras eis seruamus, sed etiam accentus, ut calefácio calefácis calefácit, tepefácio tepefácis tepefácit. in secunda enim et tertia persona paenultimas acuimus, quamuis sunt breues. similiter calefio calefís calefít, tepefio tepefís tepefít finales seruant accentus |in secunda et tertia persona, quos habent in simplicibus. 

If another non-finite verb form is compounded with facio or fio, which remain unweakened, these verbs retain not only their meaning and conjugation type, but also their stresses, e.g. calefácio calefácis calefácit, tepefácio tepefácis tapefácit.  In the 2nd and 3rd person we stress the penult although they are short.  Similarly, calefio calefís calefít, tepefio tepefís tepefít  preserve in the 2nd and 3rd person the oxytone accent that they have in the simplex.