NavList:

Alexandre, 

As far as I know, most historical efforts to obtain longitude from Jupiter's Galilean satellites involved timings of their eclipses by Jupiter (rather than their occultations or disk transits). Specifically, you could time the "last speck" of light as the eclipse took place, or the "first speck" as the satellite reappeared. Such timings are pretty definite events, and it isn't hard to time them within 10 seconds or so.

However, one of the main problems in the 19th century and earlier was that these eclipse predictions were often off by a minute or two, due to imperfect allowance for the large mutual perturbations among the four moons. The first fairly accurate predictions of these events came with R. A. Sampson's Tables of the Four Great Satellites of Jupiter (1910).

Roger