20C 



THE CIVIL ENGINEER AND ARCHITECT'S JOURNAL. 



LSeptember, 



the rcffularity of motion and avoiding of slam is obtained by 

 means of a stream of I'il being made to discharge from a cylinder 

 through a hirgc or small aperture, according to the speed required. 

 Fluids being almost incompressible, the oil will not pass through 

 the aperture beyond a given rate, which is in proportion to the 

 size of the aperture and the quantity to be discharged, and the 

 power (if the cylinder the vacuum is formed in. to press it through. 

 There is nothing in the machinery employed liable to break or get 

 out of order. 



The air-spring consists of an iron box and cover a, let into the 

 floor, which contains a vertical axle b, supported at bottom in a 

 hollow cup c, and furnished at the toj) end which projects above 

 the floor with a shoulder and lever hinge p, for carrying the door 

 on this axle; and within the box is fastened a horizontal wheel n, 

 which is toothed upon a portion of its circumference. On each 

 side of this wheel is a rack k, attaclied to a piston f, which is 

 made to fit tightly into a cylinder g, by a cap leather h. In the 

 under side of the cylinder is a valve k, communicating with the 

 outside; in the bottom of the cylinder is another valve l, commu- 

 nicating with an exhausted chamber. On each side of the racks are 

 guides M, for the piston. 



Fid 1 



The teeth of the wheel are made to take in either of the toothed 

 .1.-0 .... t1.„ ,1„,.. „,. „„*„ :„ ,1 - 11 ^„ »i . 1 11 _ 



acks 



as the door or gate is opened one way or other, so that the 

 .11 i„ .1 1 ii.- _-_T _ 1 i_ • m behind 



piston will be drawn along the cylinder, leaving a vacuui 

 at a uniform and regular degree of resistance until the door is 

 released, when the unbalanced pressure of air upon the face of the 

 piston v,-ill cause the door to resume its original position. 



riG- 2 



iiiii I 



VEETICAL SECTION. 



FIG 3 



The use of the valve K, communicating with the outside of the 

 cylinder is that, in case of a leakage of air behind the piston, it shall 

 be driven by the return of the piston tlu-ough it to the outside. 

 The use of the exhausted chamber and valve l, communicating 

 with it is, that a small portion of the 

 leakage air or oil w hich cannot be dis- 

 charged through the valve k, leading 

 outwards, escapes into the exhausted 

 chamber, which allows the piston to 

 get to the bottom, and to bring the 

 teeth of the rack in hard contact with 

 the teeth of the wheel, and thereby 

 keep the door in its proper jilace when 

 shut; in fact, gives it a maintaining 

 power. 



The regulator is for tempering the 

 speed of the door when shutting. It 

 consists of a snuiU cylinder <i, with a 

 piston T, made to fit tightly into it by hem]) packing; in the 



VERTICAL TRANSVERSE SECTION 



piston T, is a conical valve v, opening inwards to charge the cylin- 

 der w ith oil when opening the door. This valve closes wheii the 

 door begins to shut. At the end of the cylinder q, is another 

 valve, or what is commonly called a cock, which regulates the 

 discharge of the oil which passed into the cylinder during the 

 opening of the door. According to the size of the aperture z, in 

 the cock, so is the time it takes to discharge the oil. and so is the 

 speed of the door in resuming its position when shut, which com- 

 pletely prevents the motion increasing beyond what is wanted 

 and avoids slamming. 



The box requires to be filled with lard or sperm oil up to the 

 dotted line o, to seal the piston and keep the whole lubricated. 



These hinges have been used for some of the public establish- 

 ments in Edinburgh with success. 



ELASTICITY OP CAST-IBON. 



The Ili/prrliolic Law of Elasticity of Cast-iron. By Homebsham 

 Cox, H.A. Jesus College, Cambridge. 



The object of this paper was to show that the extension and 

 corresponding tensile force of a cast-iron rod are related to each 

 nearly as the ordinates of a hyperbola. That the tension and 

 extension are not directly proportional, but that there exists what 

 is termed a defect of elasticity was shown by Leibnitz, James Ber- 

 nouilli, and others. 



The real law of elasticity of any material can be ascertained 

 only by direct experiments, and dift'ers slightly even for two difl^er- 

 ent specimens of the same material. All, therefore, that can be 

 done in expressing the law by a formula is to represent the average 

 of several experiments. The results of a set of experiments can 

 be represented with any required degree of accuracy by a formula 

 expressing the weight by ascending integral powers of the corre- 

 sponding extension. The ordinary law stops at the first term of 

 the series; and the modification which most readily suggests itself 

 is to extend the series to the second term; so that if e be the lon- 

 gitudinal extension of a cuniform rod, w the weight producing it, 

 and A and B empirical constants, 



w ^ Ae — Be- (1) 



From the experiments recorded in the Report (I84-9) of the 

 Commission "appointed to inquire into the application of Iron to 

 Railway Structures," it is manifest that the for' ula (1) adopted in 

 the Report is suhjeet to unavoidable inaccuracic.-:. Eight forniulse 

 are given for extension of diff'erent kinds of iron; and it is obsen- 

 able in each case, without exception, that at least one-half, and 

 generally more, of all the results of each set come together in the 

 middle of the series with the errors in excess, and are preceded 

 and followed by results in which the errors are in defect. The 

 general character of the errors is therefore this — they are at first 

 negative, then positive, and increasing in magnitude up to some 

 term near the middle of the series. They then decrease till they 

 become negative again. It may be shown by simple algebraical 

 reasoning that the error may be nearly expressed by the first, 

 second, and third powers of the weight; and this expression, added 

 to the original formula, gives a cubic formula which is more cor- 

 rect than either the quadratic or bi-quadratic formula as obtained 

 in the Report. 



All these forniulse lead to very complicated results in their ma- 

 thematical applications. That, however, which is here proposed, 

 possesses the advantage of far greater simplicity combined with 

 accui'acy greatly exceeding that of the quadratic formula. 



Fig. 1. Fig. 2. 



If f^ be the extension of a rod produced by a stretching weight 

 «', it will be found by examination of the experiments that the 



