The boulders meet the erratic category in that the underlying bedrock is migmatite of the Chelan Complex - highly metamorphosed igneous rocks. Further up the canyon the underlying bedrock is well exposed.
Road cut into the Chelan Complex migmatite
During past drives on the road I had kept an eye out for the contact between the Columbia River Basalt Group and the underlying metamorphic rocks. The place where the flood basalt covered over the old soil and bedrock 15 million years ago. But there is no exposure of that contact here.
Near the top of the canyon the road cuts through a unit of silts and fine sands which in the past I assumed was associated with some sort of glacial outwash or ice margin lake as the silt and sand were capped with glacial till. That interpreation had been made by a quick glance out the car window without stopping.
This time I stopped and took a closer look.
Slightly tilted silts and sands
First thing to note is that these silts and sands are tilted off the horizontal. A bit of a surprise for young glacial units and cause to suspect these are not glacial sediments. The silts and sands are capped with glacial till.
Silt and sands capped by basalt-rich till.
Note the till has filled in an area where the silts and sands were eroded by the ice.
On the downhill side of the road cut is an exposure of basalt with the lake sediments overlying the basalt.
Columbia River Basalt Group underlying the lake sediment
A close look at the contact shows the lake sediment has been cooked by the basalt.
These old lake sediments are not a glacial ice margin lake. They are older than the basalt. The lava flow flowed into a lake and invaded the sediments as an invasive lava flow pushing the sediments aside with the heat of the lava then altering the silts and sands in close contact. The silts and sands may be correlated with the Ellensburg Formation - alluvial sediments deposited between basalt flow events during the time of the Columbia River Flood Basalt eruptions.