In Book 2, lines 433-508, of his Libri Evangeliorum Quattuor, Juvencus places a long discourse of Christ in which he gives His disciples their instructions. Line 462, in a passage in which Christ tells the disciples that they will be persecuted and punished for His sake, includes an excellent example of sound corresponding to sense. The subject matter of the passage is harsh, as Christ mentions specific torments they will suffer, and the ugly sound of the line reflects this well. Here is the couplet 462-3:
Vos flagris vinclisque feris durisque tyrranis
frendens urgebit pro me violentia saecli.
Both lines have a rather heavy feel. The first scans -----^^---^^-x, while the second scans -------^^-^^-x. In 462, every word ends with an 's'-sound (if we discount the copulative -que endings), and the second through sixth words all end in long -is. Moreover, in flagris and feris, the pulse and ictus clash, so that the -is sounds receive a special sort of emphasis (I'm assuming that pulse and accent coincide for vinclisque and durisque due to an accent shift to the penultimate caused by non-elided -que--which reminds me: in Golden Latin Artistry Wilkinson states, I believe, that the question of whether accent shifts in a word containing an elided -que is undecided. That book was pubished in 1963; does anyone know what the current status of that question is?).
Perhaps I'm being fanciful, but I also detect in these slithering 's'-sounds a hint of the serpent, popular since the Garden of Eden as a figure for Satan and an appropriate one to evoke here in the context of the evil punishment of the disciples for the sake of the righteous Christ. If this hint is present, another interesting layer of meaning is added as well by the fact that only four lines earlier Christ had instructed His disciples to be as cunning as serpents (458):
sed vos arguto serpentum corde vigete...