a very much larger body of water.  Although we were here at the beginning of summer, the wind that came down from the mountain at night was very cold.  Our minimum thermometer registered 22 deg.  F. near the banks of the lake at night.  Nevertheless, there was only a very thin film of ice on the borders of the lake in the morning, and except in the most shallow bays there was no ice visible far from the bank.  The temperature of the water at 10:00 A.M. near the shore, and ten inches below the surface, was 61 deg.  F., while farther out it was three or four degrees warmer.  By noon the temperature of the water half a mile from shore was 67.5 deg.  F. Shortly after noon a strong wind came up from the coast, stirring up the shallow water and cooling it.  Soon afterwards the temperature of the water began to fall, and, although the hot sun was shining brightly almost directly overhead, it went down to 65 deg. by 2:30 P.M.

The water of the lake is brackish, yet we were able to make our camps on the banks of small streams of sweet water, although in each case near the shore of the lake.  A specimen of the water, taken near the shore, was brought back to New Haven and analyzed by Dr. George S. Jamieson of the Sheffield Scientific School.  He found that it contained small quantities of silica, iron phosphate, magnesium carbonate, calcium carbonate, calcium sulphate, potassium nitrate, potassium sulphate, sodium borate, sodium sulphate, and a considerable quantity of sodium chloride.  Parinacochas water contains more carbonate and potassium than that of the Atlantic Ocean or the Great Salt Lake.  As compared with the salinity of typical “salt” waters, that of Lake Parinacochas occupies an intermediate position, containing more than Lake Koko-Nor, less than that of the Atlantic, and only one twentieth the salinity of the Great Salt Lake.

When we moved to our second camp the Tejada brothers preferred to let their mules rest in the Puyusca Valley, where there was excellent alfalfa forage.  The arrieros engaged at their own expense a pack train which consisted chiefly of Parinacochas burros.  It is the custom hereabouts to enclose the packs in large-meshed nets made of rawhide which are then fastened to the pack animal by a surcingle.  The Indians who came with the burro train were pleasant-faced, sturdy fellows, dressed in “store clothes” and straw hats.  Their burros were as cantankerous as donkeys can be, never fractious or flighty, but stubbornly resisting, step by step, every effort to haul them near the loads.

Our second camp was near the village of Incahuasi, “the house of the Inca,” at the northwestern corner of the basin.  Raimondi visited it in 1863.  The representative of the owner of Parinacochas occupies one of the houses.  The other buildings are used only during the third week in August, at the time of the annual fair.  In the now deserted plaza were many low stone rectangles partly covered with adobe and ready to be converted into booths.  The plaza was surrounded by long, thatched buildings of adobe and stone, mostly of rough ashlars.  A few ashlars showed signs of having been carefully dressed by ancient stonemasons.  Some loose ashlars weighed half a ton and had baffled the attempts of modern builders.