Recent studies indicate three problems with the concept of continental drift as an incidental corollary of plate movement: (1) Slab pull can not drive plates with continental leading edges, (2) There is no low velocity zone under shields, and (3) Continents have ''roots'' 400 to 700 km deep. These problems imply that if continental drift occurs, it must use mechanisms not now understood, or that it may not occur at all, plate movement being confined to ocean basins.