impinge. These thermocouples are laid in fire-clay tubes in the joints, with the points one-sixteenth inch back of the face of the wall, these ends being protected from the direct heat by a coating of fire clay. Readings on all ten couples are taken at intervals of ten minutes by means of an Englehardt galvanometer.

Five thermometers are hung on the back of the wall for the purpose of determining the rate of heat transmission. The bulbs are held against the blocks by covering them with fire clay, which prevents the cooling effects of air currents.

After the door has been subjected to the heat for two hours the gas is turned off and the door is removed from the chamber. In some of the tests the blocks are allowed to cool slowly in air, and in others they are cooled suddenly by water.

The quenching test consists in directing a stream of water against the hot blocks. The door can be run out of the furnace and the water applied in about thirty or forty seconds after the gas has been turned off. The stream is played for five minutes through a 3-inch nozzle at a distance of 20 feet from the door and at a pressure of 50 pounds per square inch. (See Pl. XXIV, B.)

After the blocks tested in the recent series were sufficiently cool they were removed to the Armour Institute, and, by the courtesy of Prof. R. Burnham, in charge, they were tested for transverse and compressive strength, using the same methods as those used at the laboratories in St. Louis.

Form Q is used for recording the results of the fire tests.

Max. temperature.

Water test....... minimum at.. pound per square-inch pressure.

PERMEABILITY SECTION.

Outline of investigations. The programme for the permeability investigations includes a preliminary series for the purpose of determining the best method of procedure. The subsequent series covers tests on cement mortars and concretes-those treated with waterproofing coatings and those in which different kinds of waterproofing compounds have been incorporated. Different mixtures, consistencies, and thicknesses are used and the specimens are tested at different ages. Tests will be made on specimens treated with

all the available waterproofings, including powders and liquids for incorporation in the mixture and also for exterior application.

In addition, 2-inch cubes of mortar and concrete will be made and tested for porosity in order to find the relation between porosity and permeability.

Apparatus.-The apparatus for these tests is shown in Pl. XXV and fig. 8.

The arrangement for holding the test piece and the can for catching the water that passes through it are shown in Pl. XXV, A. The piece of apparatus shown at

the right is placed on top. and that next is placed on

be

the bottom of the test piece, with rubber washers tween each cap and the test piece. A test piece is shown just to the right of the can. A cross section of the apparatus for holding the test piece, with a test piece in it, is shown in fig. 8. The water is applied to the top of the test piece through the pipe C; the distance Y is the height of the test piece, and A shows the position of the rubber washers at the top and at the bottom of the test piece. The water that passes through the test piece is measured, being caught in the can B attached to the bottom of

the casting that holds the test piece.

A view of the room used FIG. 8.-Cross section of apparatus for holding permeability test pieces. for this work, with two per

meability tests under way, is shown in Pl. XXV, B. Twelve tests can be carried on at once. A diagram of the apparatus, with pipe connections, is shown in fig. 9.

The water passes through the filter A and into the tank B, from which it is fed to the test pieces attached at the pipe connections. Tank D is connected to the air compressor on the floor below and is used to equalize the variable pressure of the compressor. The air is introduced above the water tank B at the point E. The pressure

of the water is read at gage F, and of the air at gage G. The molds for the test pieces are short sections of iron pipe of 7 inches internal diameter and of varying lengths.

Method. The mortars and concretes used in the permeability tests are mixed on a glass-top table. The material is firmly pressed into the mold by hand, the object being to do as little tamping as possible in order to prevent the flushing of cement to the surface. The top of the test piece is smoothed off level with the top of the mold by the use of a 10-inch flat trowel, and is then scraped rough at the surface so that it will have the same characteristics as the interior.

Section A-B

FIG. 9.-Diagram of permeability apparatus and connections.

The test pieces are stored in a moist room of the block section until ready for testing.

The test pieces are placed in water for forty-eight hours before being tested. Annular spaces at the outer edges of both the top and bottom surfaces are painted with a rubber waterproofing paint, leaving a circular area 5 inches in diameter in the center of the specimen in its original condition. Rubber washers are placed over the waterproofing and the specimen is securely fastened in the holder. The apparatus is then attached to the unions (C, fig. 8) and the water is turned on. In the present series a pressure of 20 pounds per square inch is maintained. Higher pressures are to be used in the subsequent series.