if yer speaks are truly bi-ampable, and ya *do* have two identical amps, the best ting to do is yust try the 3 different methods - bridged, wertical, or horizontal bi-amping. it's easy to tell if yer speeks w/two sets of binding-posts are *really* biamp-able: yust hook one amp to only one set of binding-posts w/o the yumper in place, and only that driver should play - the other should be silent. (but, it's hard for me to imagine why a speaker mfr wood have two sets of binding-posts, if all drivers play when only one is used, w/o a yumper).
i have had good luck horizontally biamping w/an electrocompaniet aw100 driving the woofs & an aw75 driving the tweets of my monitors. but, yust this week, i sent my aw100 to a guy who wanted to trade his aw75 for a bit more power. his aw75, w/serial # only 5 digits away from my aw75, should give me the perfect opportunity to try bridging & wertical bi-amping, as well as horizontal bi-amping. i also have a pair of gnu aw60's, bought from europe, waiting for me to conwert from 220 to 120v. (i had given up on finding an aw75, & the trade offer happened shortly after i bought the aw60's!). so, i may also try bridging all 4, so each driver has a mono amp. but, this may really be over-kill in my system, so i may end up using 'em in another system, trying to sell 'em & use the money for a pair of melos mono-toob amps, or...??? ;~)
re: ral's suggestion of using an active x-over, i'd recommend this *only* if ewe can completely bypass the speakers' internal passive x-over - i wood *not* use an active x-over *and* the the speek's passive x-over - this is sure to muck-up the signal, imho. and, an amp run to one set of binding-posts of a bi-amped speaker only sees the the load above or below the speakers' passive x-over point, anyways...
good luck, doug s.