Reverse Inclination basing is a method of housing
ICBM silos so as to (theoretically) provide them with maximal protection from attack by other
ICBMs, and was bandied about as one of the potential means of situating the
MX missile in the early 1980s. This concept was brought to you by the
Reagan administration, the color
puce, and the letters
wtf?!.
By 1980, the science and engineering of building ICBM silos had advanced pretty far along. Silos, in addition to being embedded in thousands of tons of ferroconcrete for strength, had their sensitive bits riding on massive springs to survive ground shock. They had their electrical systems shielded and earthed to avoid EMP. They had massive concrete doors to ride out nearby atomic detonations. The key word, however, is 'nearby.' While it was fairly certain that an airburst wouldn't harm, say, a modern Minuteman III silo to the point of it being unable to fire, a ground burst was another story. Even if it didn't penetrate the literally yards of shielding over the missile, there was a good chance that it might simply toss up enough earth in the detonation to bury the opening of the silo such that it wouldn't be able to open and/or fire.
This is not what you want to hear, if your nation's defense rests on said missile silos' ability to ride out and survive a first strike. Around this time (1980) the MX missile was having all manner of trouble in Congress regarding its cost, its performance (some considered it too accurate) and its basing modes. In addition to the problem above involving ground bursts, massively hardened silos are expensive. One (cheaper) method of survival that was examined was reverse inclination basing.
Inbound ballistic missiles follow a fairly predictable general flight path. After all, missiles aimed at the U.S. from the Soviet Union would likely fly over the Arctic and attack targets from the north. As they are following a slightly modified ballistic arc, there will always be some southerly component to their flight path - in some cases, quite a bit of it. Therefore, if one could find a large enough and steep enough mountain or large hillside, with a southward slope, it might be possible to emplace missile silos on this southern slope such that an inbound warhead would be unable to come down steeply enough to detonate in a ground burst - which would mean that the silo would be effectively invulnerable.
While this is a nice idea, it falls short in a few ways. For one, unless you had a really tall mountain, it would be difficult to gain all that much distance from the detonation (the geometry is left as an exercise for the reader, and remember that atomic explosions suck at even long distances). In addition, while it might be possible to avoid the silo taking severe damage directly, it would be really easy to drop tons of mountain down on top of it - anything steep enough for this not to be a problem would be damned impractical to bore a silo into and base missiles in. The downslope side of the silo would also naturally suffer fairly poor shielding characteristics, rendering it (perhaps) vulnerable enough that even said nearby airburst would do the trick. Finally, SLBMs made the entire argument groundless, since they could approach the target from non-Northern (and, indeed, variable) headings. As the U.S. realized that the Trident II missile was accurate enough to be used as a first strike weapon, it had to assume that (eventually, at least) Soviet SLBMs would achieve similar performance.