With his mind kindling to the idea Bryant rode northward next morning
along the base of the mountains, studying the hillsides where a canal
naturally should run, all the way up to the Pinas River. Afterward he
reconnoitered the mesa, hitting at last on a slight elevation, hardly
to be called a ridge, that projected from a hillside a mile below
Bartolo and curved in a gentle crescent for about three miles from the
range of mountains down the mesa, again bending in toward the hills
close to the north line of the Perro Creek ranch.

Next, he absented himself for a week at the state capital, where he
industriously studied the water and land records pertaining to the
district. When he returned, he brought with him a surveying instrument
and a boy for helper. He pitched a tent out of sight in a hollow at
the foot of a hill, worked early and late running his lines,
establishing a dam site, and surveying the river bottom near the mouth
of Pinas Cañon, and remained practically unseen except by a few
incurious Mexicans. His instrument proved the correctness of his
conclusion regarding the crescent-shaped elevation as a practical
grade for a canal, which though necessitating a longer course would
nevertheless immensely lessen the time, expense, and difficulties of
digging when compared with a line along the mountains' flanks with its
danger of washouts and earth slides. Nor did he stop there. He made
rapid but reliable topographical measurements, on a general scale, of
the mesa for five miles out from the mountains, between Bartolo and
Perro Creek, locating among other things a large depression in the
plain, three miles southwest of the town, which might by diking be
converted into a flood water reservoir. Then he folded his tent and
again disappeared for a week. When, finally, he rode to Stevenson's
ranch house that hot July afternoon and made a trade for the five