digitōs meōs praerōdō — I chew my nails (literally, ‘I gnaw at the tips of my fingers.’
This second one is conjecture. The phrase is sourced in Plautus, Pseudolus, where the image is of guests literally gnawing at their fingers because they are enjoying a feast so mindlessly that they lose track of where the ham ends and the hands begin. That said, ‘chewing the tips of one’s fingers’ could easily fit with the image of gnawing at one’s nails—don’t you think?
There’s nothing in the L&S entry for unguis to settle the case, but here’s the L&S entry for praerōdō—http://goo.gl/pYh68p
I trust that if you’re reading this, you understand that many English words ‘are derived from’ Latin counterparts, though we can further distinguish this by stating that there are two varieties of derivation: direct and indirect.
Direct Latin derivatives (for instance ‘fact’ from the Latin factum) are more or less coequal adoptions, whereas indirect Latin derivatives (such as ‘feat’ from the French ‘fait’ from the Latin factum) feature a few sound shifts which echo the modifications of the mediating language. Another example: from dāta we have both ‘data’ (direct) and ‘date’ (indirect, through Old French ‘date’).
[A more interesting etymology, while we’re at at it: the English homograph ‘date’ (the fruit) is from the Old French ‘datte’ from the Old Provençal ‘datil’ from the Latin dactylus (the same fruit), so named because it resembled the human finger and/or because this word resembled the Semitic names for date palm: deqel/daqal, etc, which have nothing to do with fingers.]
If you’d like to rabbit-trail even further, here’s a post offering the Latin names of the five fingers, including the pinky finger, whose name made my day: (http://www.omniglot.com/blog/?p=442)
If anyone knows of a Latin derivative which is indirect but not mediated by French or Old French, I would love to see it in the comments below.