here where a pdf link is available.
Readers of this blog will recall my last posting about a proposal made by Dr. Jim Sears (University of Montana) at the Annual Meeting of the Geological Society of America in Denver last week. In it he proposed that the ancestral Colorado River might have flowed west through an early incarnation of the Grand Canyon to an area near Nevada, then turned northward toward Idaho and Montana, eventually spilling into the Saskatchewan arm of the pre-Pleistocene (before 2.5 million years ago) Bell River system. The Bell River was an Amazon-scale river system that drained about one-third of the North American continent before glaciers in the last 2.5 Ma reorganized the drainage system.
This arm of the Bell River thus crossed the present-day Snake River Plain and the Idaho/Montana Continental Divide to connect with the Saskatchewan arm of the Bell River. Note a greatly diminished Mississippi River in the pre-Pleistocene United States.
Of course, this idea is not a slam dunk immediately. Further testing of the idea is necessary. But it does dovetail with other proposals by Karl Karlstrom and Brian Wernicke that a paleocanyon in eastern Grand Canyon might have been present during the Miocene.