TRANSACTIONS OF SECTION G. 775 



the deptli. For this particular deflection it is shown that the factor of safety 

 multiplied by 9 p;ives the critical point. 



If the deflection in percentages of the span are calculated a series of curves of 

 deflection can be plotted. At 1 ton per squai-e inch and 1 per cent, of deflection 

 the critical point is 576, the depth at 2 per cent, of deflection, the critical point is 

 2S8 times the depth, or 9 x 32. As at 1 ton per s(iuare inch, 32 is the factor 

 of safety. 



The values are tabulated and shown graphically by diagrams. 



Returning to the special object of the paper, the selecting of rolled steel joists 

 for tireproof doors, the principal step is to determine the pitch or spacing apart 

 of the joists. 



These pitches are tabulated for various sections of joists for the loads of 

 1 cwt. and Ij cwt. per square foot of floor. A formula is deduced for a loading 



of 1 cwt. per square foot: p^"— — , and for any other loading /j^.-C., where .f 



is the cwts. of load per square foot, p = pitch in feet, L = span in feet, and W ^ dis- 

 tributed load in tons, the rolled steel joist will carry safely. 



A final result is that the pitch varies inversely as the square of the span when 

 the stress per square inch is considered, i.e., up to the critical point, and varies 

 inversely as the cube of the span when the deflection is considered or beyond the 

 critical point. 



7. On Alternating Air Currents in Churches and Public Buildings, 

 Bij J. W. Thomas, F.I.C, F.C.S. 



When the temperature of the air outside is 35° F. or less, the exit space for 

 foul air in a great number of churches and public buildings is too large to keep 

 back the extra pressure outside, and cold air enters the top of the building at the 

 points of least friction and resistance — the large openings generally. In hio-h 

 buildings the cold air currents, or down-draughts, are followed by hot and 

 oppressive waves of air, after which the air becomes motionless and stagnant for 

 some seconds. 



Some years ago the author experimented in a large pubhc hall, and found that 

 these hot and cold experiences were due to alternating air currents in the buildino-. 

 Taking the point of the least internal pressure as the first observation, it took 

 about half a minute to reach the point of highest internal pressure, and rather less 

 than half a minute afterwards to reach the point of least pressure arfaiu. The 

 first five seconds alter the least internal pressure was reached there was a gradual 

 rise, followed by double such an interval of more rapid increase ; then there were 

 a few seconds of lesser increase, followed by a lengthened period, durinn- which 

 the pressure-recording instrument remained almost steady. When the reduction 

 of internal pressure began much cold air still descended, and there were ten or 

 more seconds during which the reduction was gradual ; then, for about half that 

 period, a very rapid decrease occurred, followed by several seconds when the 

 instrument was steady and almost stationary at the point of least pressure. 



The strangest fact in the results obtained was that, owing to the elasticity of 

 the ail', its density (32° F. outside), and the velocity obtained by falling about 

 60 feet, the pressure increased internally until it actualhj exceeded the pressure 

 outside for a few seconds, then decreased and increased alternately. 



Since then experiments in high churches and buildings have given similar 

 results. An anemometer held in a narrow opening in a doonvay leadino- to a 

 church turned rapidly inwards, indicating an up-current ; then it stopped and 

 subsequently reversed, showing that the pressure in the building was actin"' 

 outwards. 



Air inlets to hot-water pipes under the floor of a building are influenced bv 

 alternating air currents at their highest pressure, and when the period of greatest 

 upward movement occurs, such a deluge of cold air passes inwards that the inlets 

 Uuvc to be shut. 



Altercating air cui'rent&j therefore, greatly iiiajjak tbe veutilatiou of buildings. 



3 E 2 



