Housman reads assueta euolitans; the former word is a conjecture of his own, the latter a conjecture of Ellis, which I think he would have ignored if the relevant fascicle of the Thesaurus had been available to show that euolitare occurs once in Columella and then (if at all) not before the sixth century. If assueto is sound, mundi must be changed to mundo (an isolated variant) or to another noun. Bentley read mundo, and this may well be the right solution: the eagle carries thunderbolts to the sky, “cui scilicet per diuturnas operas assueuerat”. Shackleton Bailey (1979, 162) emends to nisu (on the theory that this fell out after -mina and was replaced by the stopgap mundo). If emendation were required, I suggest that motu would be palaeographically more satisfactory; the assumed process of corruption is well illustrated by 3.82, motu GL: modum M: mundo LV.