by the United States Department of Agriculture in Bulletin 403, and we, therefore, present certain extracts from that bulletin with some additional notes.

On account of the variations in the size of the sand grains and in the unfilled spaces between the particles of sand, stone and gravel, the quantities of concrete made, according to certain proportions may vary. For the same reasons the quantity of water may or may not be sufficient to make the concrete wet enough. Such matters, with the experience gained through making a few batches, may be easily adjusted. Water should always be measured by the bucket to have uniform results.

Different pockets of sand and gravel and different crusher-run rock vary in size and consequently in the unfilled spaces or voids between the grains or pieces. This variation, in theory, calls for varying quantities of cement, but the methods of determining the exact amount of these unfilled spaces are so complicated and so open to mistakes that with cement at its present price it is cheapest in the long run to adopt proportions or mixtures, which are rich enough to guarantee a well-made strong post. If unscreened bank run gravel is decided upon it should be used in the proportion of one part of cement to four parts of gravel. For crushed rock or screened gravel (which is much better than bank run gravel) the concrete should be used in the proportion of one part of cement, two parts sand and four parts of rock or gravel. Also one part cement to four parts chats makes good concrete.

All measurements should be made with the material poured loosely into the measuring box and the box when full should be leveled smooth.

The amount of moisture in sand, gravel and stone varies so much with weather conditions that the quantity of water for a cubic foot of concrete cannot be fixed exactly. During the mixing of the cement with the sand and rock sufficient water should be used for the concrete to be wet enough when the mixing is complete to tremble under a blow from the shovel. This amount of water causes a rich mortar to flow to the outside of the post and insures a smooth finish.

Bulletin No. 384 of the United States Department of Agriculture affirms that experiments conducted at the Colorado Experiment Station show that poured posts were a little over 25 per cent. stronger than tamped posts of the same size, mixture and reinforcement.

MIXING CRUSHED ROCK OR SCREENED GRAVEL,

In general, where crushed rock or screened gravel is used, the full amount of sand is spread out upon the board and upon it the necessary cement is evenly distributed. The whole is turned dry until the cement no longer shows in streaks, and the color of the batch is uniform. The mixture is then spread out flat, just as the sand was, and upon it the crushed rock or screened gravel is distributed evenly. Three-fourths of the water required is added and the mixing is continued. In dry hot weather it is a good plan to throw water on the pile of crushed rock before mixing.

OILING THE MOLDS.

Concrete has a erlency to stick to either steel or wood. In order to yield a smooth finish to the post, it is customary to give the inside of the molds a coating of il. Soft soap or crude oils used sparingly serve the purpose well. Too much oil will destroy the setting qualities of the cement and will give a ice roughened with pockmarks. A small amount of oil, poured into a pailof water and applied with a mop or stiff broom in scrubbing out the mols after they have been used five or ten times, or as often as necessary, vill prevent the concrete from sticking.

MOLDING.

After the molds, whin, as a rule, lie flat, but not always so, have been oiled or soaked, the oncrete should be placed in them at once. If for any reason the concret stands thirty minutes after mixing, it should be thrown away and a new atch mixed, for cement, if it has once partially set, makes weak, dangerou concrete, even though it is retempered by turning or adding water. f wooden molds are used, they should be well soaked in water so tht the concrete post will not cause them to swell and thus crack the pos The wind, the hot sunshine and the frost are destructive to concrete; nd while it is an easy matter to protect the post from the first two agences, it is best to do no concrete work when the thermometer is below thefreezing point, unless inside a building.

RINFORCEMENT.

Reinforcing should be plaed near the outside wall, where it is reasonable to expect that crack will open. For reinforcement, metal slightly rusted is as good as retal that is not rusted, if not better. In placing the reinforcing rods in position it is a wise precaution to bend them back at the ends. This tkes only a little more time, but, if it is done, the reinforcement must be 2 in. longer to allow the metal to be turned back I in. at each end o the post. New barbed wire should not be purchased for reinforcement, for while it costs more than plain reinforcement, the bond between it and the concrete is no stronger than between smooth wire and concrete. The danger section, or the point where posts are liable to break, is at the surface of the ground. For fences for lots and other places where posts may be subject to rubbing and crowding, short extra reinforcing pieces 2 ft. long are sometimes placed in the post to lap this danger section. The corner of triangular-shaped posts, which is not nearest the fence wir should, in theory, have heavier reinforcement than either of the other corners. If rods three-sixteenths of an inch in diameter, No. 8 wire, a two twisted No. 12 wires, are used in the other corners, a rod one-fourth of an inch in diameter, or two or three twisted No. 12 wires, shoud be placed in the threatened corner.

CURNG.

It is a great mistake to believe that, when the molding is done, a concrete post is finished. The qualty of the post must be determined

by curing. The green post should be left in the nold until thoroughly hardened; that is, usually for two or three days

During the first two days of the life of a ost it must be kept wet and covered with canvas, burlap, carpet, or ay lean material. Sand will serve after the concrete has become had, but manure will stain green concrete and otherwise affect it. The sprinkling should be continued up to the eighth day. After the tenth ay, if the space is needed, the post may, with care, be placed on end in the same manner that wooden fence rails were formerly piled. A drop of only 6 in. often breaks a green post. The jar in hauling to he field over rough, frozen roads or in a wagon bed with a very unevn bottom has seriously injured posts which were not well seasoned. Oncrete posts gain rapidly in strength for the period of one year; they sould, therefore, be made as long as possible before it is necessary to se'them in the fence. No post should be used until it is at least 2 month old, and, to meet any contingency, a supply of well-seasoned posts sould be kept on hand.

METHODS OF ATTACHIG WIRE.

There are numerous methods of attaning wire fencing to concrete posts. Some makers place staples or wir loops in the green concrete: others make holes in the posts. The forrer method is not desirable, because the fastener cannot be located exacly where the wire of the fenc ing will come when the post is set in the ground; then, too, the fastener will eventually rust or break off and wll thus injure the looks of the post. On the other hand, holes through the posts weaken them and therefore this method is, in general, unstisfactory.

The simplest, easiest and cheapest ray of fastening a wire fence to a concrete post is by encircling the pot with a wire one size less than the corresponding wire in the fence poper, and by twisting this wire around the strand of the fence. This is done in two ways. The fastening wire is placed around the post, twisted upon itself and then to the fence wire; or one end of the fastening wir is twisted around the fence wire, and the free end is then carried around the post and twisted on the other side to the same wire. The latter method is known as "the Western Union twist." Either plan is good, but care must be taken to draw the fastening wire tight, or else stock trying to get through the fence may raise or crush down the fencing with their heads. If any trouble is experienced, the post should be rouglened at the fastening point with a cold chisel.

Expansion and contraction of the fence, due to heat and cold, are cared for by the tension curves of "kinks" in the woven wire fencing. and no fear may be felt in drawing the fastening wires as tight as necessary. Wooden nailing strips should never be embedded in the posts, moisture will swell the wood and crack the concrete.

for

TAPERED POSTS.

In the early stages of concrete post making it was the opinion of some, and yet remains 5, that posts need not be tapered from base to top, but from the expetence and testimony of those who have had a large practical experene in the matter, it has now been adopted as almost unanimous jugmnt that taper is necessary for many good and essential reasons.

It has been definitely emonstrated that it prevents heaving in winter, almost entirely if not abslutely. A fence post acts as a cantilever and the post is therefore largst toward the ground line, where it is the most severely stressed. It also reduces the amount of concrete required, lessening the quantity owed top of post where the full section, as at the base, is not required. The tapered post lends itself most naturally to the fastening of the fening wire by means of the loop around the post with the Western nia twist; the taper prevents displacement of the loop.

ONCLUSIONS.

(1) Concrete posts are practical, economical and a suitable substitute for wood.

(2) Reinforcement shoul be placed as near to surface of post as possible; 1⁄2 in. from surface i best location.

(3) Post should taper frm base to top.

(4) Post should not be lss than 51⁄2 in. at base and 4 in. at top. (5) Concrete mixture shuld consist of one part cement to four of run of pit gravel; or one prt cement, two parts sand and four parts crushed rock or screened grael. Gravel or crushed rock not to be smaller than 4 in. in size, nor larger than 1⁄2 in. Concrete should be of a quaking consistency.

(6) Molds should have a jogger or vibratory motion while concrete is being introduced to compact it and smooth up surface of post. (7) Posts should not be made out of doors in freezing weather. They should not be exposed to an, and should be sprinkled with water the first eight or ten days after bing made to aid curing.

(8) Molds should be carefuly oiled or soaped to provide a smooth finish and to prevent concrete sticing to mold.

(9) Posts should be cured br not less than 90 days before being set or shipped.

(10) Posts should be carefuly handled and be packed in straw, sawdust or other suitable material for shipment.

RECOMMENDATIONS.

(1) We recommend the adoption of Conclusions 1 to 10 relative to concrete fence posts.

(2) We recommend that the bregoing conclusions be substituted for those in the 1911 Manual of Recommended Practice, page 210, upon the subject of concrete fence posts.

(3) We recommend to the Board of Direcion that the Committee be directed to endeavor to have a series of tests conlucted to determine the comparative strengths of the different form of concrete fence posts and method of reinforcing, with a view to ascetaining the most suitable form. These tests to be performed, if possibl, at one or more of the university or government laboratories.

(2) TRACK CONSTRUCTION AND FLANGEWAYS AT PAVED STREET CROSSINGS AND IN PAVED STREETS.

This subject has been under consideratin for several years past, and your Committee has studied the practic in European countries as well as our own. It would appear as if th street and interurban rail lines have reached a higher state of development in this respect, both at home and abroad, than the steam lines. It s but natural that it should have been so, because of the greater extent which they have been confronted with the problem.

It is probable that the most advanced pe of construction will suggest and determine the best design for stear road work. They have been in such intimate contact with the difficultes encountered in track construction in municipalities, where the requiements for street paving were very rigid, that they have now achieved type of construction that is giving most satisfactory results for their ind of traffic. We submit below brief comments upon important featres of such track construction.

DRAINAGE.

The feature of good drainage, so esential to first class track construction, appears to have received abudant attention from street railway lines, and more than steam roads sem to have given it.

If the track superstructure is built n a concrete bed, the infiltration of water, which lies about the concret, soon causes it to disintegrate under the constant impact of wheels pssing over it, particularly at the rail joints, and soon reduces the concrte at such points to the consistency of mud, making track renewal oon necessary. If on the other hand ballast alone is used for the subtructure of track, the natural bed beneath it soon becomes saturated and soft, and the resulting impairment of the track is inevitable. Drainage therefore, is imperative. While satisfactory drainage provisions are mcessary in any part of track construction, there is a conspicuous obligation to provide it at road, street crossings and through paved streets.

Each situation presents a separate problem, which will have to be solved to meet the conditions of the case; therefore, no definite recomwendations can be made to cover he general proposition. It is sufficient to say that this feature of each piece of construction should receive careful and adequate treatment or the entire piece of work will prove a failure.