jeff train

A few days ago I described the basic approach I'm taking with the peninsula area. When I started laying out the path for the waterways, I quickly determined one of two options - both of which are shown in my rather rough sketch attached.
The first shows the river arranged as I've been planning - essentially the water "disappears" (somewhat gracefully, I hope) around a bend and between two low hills. One of those hills forms a "background" for the feedmill and freighthouse in Waterbury.

OPTION 1:

The area around the station/feedmill will remain a "freelance free zone" - in other words it will be prototype specific. Some of the rest of Waterbury will be freelanced as needed to best fit the space.
NOTE: The white building in the lower right is a Branchline creamery kit. The white line through the center represents a road. (hey, Rick Johnson I'm not . . .) Not sure the creamery will remain in that location.
Although this looked okay on paper, and in initial planning - looking at it now I'm thinking the river/feedmill/background hill is just a little too cramped. It just doesn't look natural. To look right a mill should have a intake and an upper pond, with a falls, and a lower pond. I have that here, but it looks too forced - and the lobe end of the peninsula would end up with a few town buildings, a street, and a bunch of trees . . . and believe me when you model New England the last thing you need is a reason to make even MORE trees . . .

OPTION 2: 

This shows my current leanings - and it offers the advantage(?) of possibly adding a second mill (with another dam/falls) in the location currently occupied by the creamery building.
More importantly, even without a second mill complex, I think the water course looks a lot more natural. The hills (represented by the stacked foamboard) can be moved slightly to the right - which will free up the area behind the feedmill and freight house.
Disadvantage - I'll have to add another bridge to the mainline along a curve in the foreground. It doesn't have to be a particularly long bridge, but bridges, especially on curves, can be a real pain. One possibility is a large culvert (or even two culverts) is one possibility. The other may be to do what Paul Dolkos suggested and simply let the river "disappear" behind some trees.
Perhaps that bane of New England modelers - all those trees - are useful additions to our bag of tricks.