TRANSACTIONS OF SECTION (J. 791 



in the B station, and tlie up train moving from C to J5 will draw air from the C 

 station into the tube and expel it in the B station. Three stations are thus 

 directly involved, and a triple-station arrangement will best fulfil the physical 

 conditions. 



Owing to the elasticity of air the outlets for expelling the vitiated atmosphere 

 must not ))e situated far from the points of greatest C!)mpression, and should begin 

 in the centre of each station underground and end in the open air above the 

 station at the surface. 



For tlie same reason tlie intakes for fresh air should be close to the points 

 where the sudden expansion of the air begins. These points are at tlie ends of 

 the tubes which the moving trains eiiter. 



Doors can be Hxed at the ends of the two tubes which the trains enter in each 

 station, and closed behind the last trains at night, so that the fresh air brought into 

 the end of the tube immediately beyond the doors will drive out the foul air into 

 the next station by natural ventilating pressure. 



Conclmion. — The providing of fresh air inlets inside the ends of the tubes 

 which the trains enter, with an outlet shaft in the centre of each station, will 

 enable the moving trains to draw in and expel enough air to keep the atmosphere 

 in good condition, even in hot summer weather. In addition to this, however, 

 there are two auxiliary aids to ventilation which enable the station masters to 

 uidke certain that the state of the atmosphere in the tubes is satisfactory. 



1. Having inlets and outlets as above, the timing of the departm-e of the up 

 and down trains as they move towards the same station will enable enormous 

 volumes of air to be driven to the surface, and a corresponding volume will be 

 drawn in. 



2. By closing the doors after some of the trains as they leave the stations fresh 

 ftir must be drawn into the tubes. 



8. Experiments in Gas Explosion. 

 By L. Bairstow and A. D. Alexander. 



9. A new Form of Mirror Extensometer. By John Morrow, M.Sc. 



Mirror extensometers have not been much used in this country. They are not 

 self-contained, and to obtain the mean extension of a specimen two .sets of 

 readings are necessary. There may be errors due to slight alterations in the 

 positions of the specimen, the telescope, or the scale. In the instriunent described 

 an attempt has been made to avoid these defects and to obtain the mean extension 

 by a single observation. 



Its special feature is the use of two mirrors placed side by side, one of which 

 is attached rigidly to the instrument and the other arranged to tilt about a 

 horizontal line when the specimen extends. The tilting mirror is carried on a 

 piece of hardened steel of diamond-shaped section. One edge of this base is pressed 

 against a piece attached to the upper pair of set-screws, and the opposite edge 

 similarly against a piece from the lower screws. The four set-screws serve to 

 attach the instrument to the specimen, and are situated at opposite ends of two 

 parallel diameters of the test-piece. Any change in the mean vertical distance 

 between them is thus a measure of the alteration in length of the centre-line of 

 the specimen, and is accompanied by a slight tilting of the mirror. A telescope is 

 so placed that the two images of a scale are seen close together, and to measure 

 extensions a convenient mark on the fixed image is taken as an index, and the 

 reading coinciding with it on the other image is noted each time the load is 

 altered. The instrument has pro\'ed very satisfactory with a magnihcation of 

 1600, when the extensions can be measured to the nearest ^jy'jjiju of "an inch. 



