The autumn weather window was not an abstraction. It was a mathematical relationship between current position, miles remaining, and the date past which mountain passes became dangerous for large herds. Norah had been studying the terrain during the morning's ride, the way she studied most things now systematically with actual attention, and she had noticed something Elias hadn't included in his plan. The east wall of the valley had a natural shelf, a raised strip of firm ground running parallel to the creek for most of the two-mile length. It was wide enough for cattle to move single file or in loose pairs.
If the herd could be directed onto the shelf rather than the creek bottom, the worst of the compression risk in the narrowest sections would be reduced and the crossing could potentially be managed in two groups instead of four. It was a reasonable observation. It may even have been a correct one. What Norah did with it was the problem. She should have taken it to Elias before the drive entered the valley.
She knew that she had three weeks of evidence, that his process for decisions included a full hearing of the available information, and that he revised his plans when the information warranted it. He had done it twice on the drive already once at the first water crossing. Once when a hand reported changed ground conditions ahead of the schedule. He was not a man who refused good information. He was a man who wanted it before the decision, not during it.
But she was riding near the valley entrance when the observation clarified fully in her mind, and the first group was already at the threshold, and she made a quick calculation that talking to Elias would take 10 minutes she wasn't sure they had, so she acted. She pulled alongside the two young hands managing the right flank of the first group, Pete. Pete, the younger man, and a hand named Garrett, she'd ridden near for most of the drive, and told them to push the cattle toward the east shelf rather than the creek bottom. She gave them the reason in two sentences. They looked at the shelf, looked at her, and nodded.
It made sense to them, and she was riding with the confidence of someone who had been making good calls for 3 weeks. They redirected the first group onto the shelf. For about 4 minutes, it worked. The shelf was firm, and the cattle moved along it with more willingness than they'd shown for the creek bottom approach. Norah was riding the edge of the shelf, watching the animals move, allowing herself the specific, quiet satisfaction of a correct observation put into effective action.