580 PROCEEDINGS OF SECTION H. 



30"^ there was no perceptible pressure on the roof. When the 

 wall is extended in the form of a parapet this sheltering effect 

 is largely enhanced, so much so that witli a parapet one-sixth of 

 the total height of the roof 70 per cent, of the pressure was re- 

 moved from a roof of 60° pitch, and one of 30° pitch actually 

 experienced a negati^'e jiressure to a slight degree. Experiments 

 were also made on the lifting effect upon a roof of a building, 

 having the two sides and one end closed, but the other open to the 

 wind, and it was found that the upward pressure was equal to the 

 pressure of the wind upon a normal plane. 



Girder Bridges. — The case of a bridge, consisting of two plate 

 girders connected by a deck placed either at the top or bottom of 

 the girders, was studied, and it was found that if the distance 

 between the girders was equal to their depth the leeward girder 

 was fully sheltered, but that if that distance was double the 

 depth one-fifth of the area of the leeward girder should be added 

 to that of the windward girder, as being exposed to the wind 

 pressure. 



Lattice Work. — A card Gin. by 8in., or 48 sq. in., was tested in 

 comparison with a similar card in which sixteen rectangular open- 

 ings had been cut, thus changing it into a grating, the area of 

 which was 27'4 sq. in., or 57 per cent, of the solid card. The 

 pressure on the grating was, however, found to be 83 per cent, of 

 that on the solid card. Placing the grating in front of the card it 

 was found that the jjressure on the latter was reduced as follows : — 



Card 6in. x 8in pressure •371bs. 



Card grating l^^in. in front " "23 " 



Sin. " " -17 " 



4iin. '' " -12 " 



6in. ^' " -11 " 



" 7iVin. " " -16 " 



9in. " " -20 " 



During the course of the experiments my assistant, ]Mr. Jas. 

 Mann, called my attention to two very curious phenomena. The 

 first of these was the shelter a small surface receives from a larger 

 one placed behind it. This was tested in numerous cases, one of 

 the most striking of Avliich was when a card 3in. square was ex- 

 posed to the wind and a 9in. diameter disc placed behind it. Without 

 the disc the small square received "ISlbs. pressure, with the disc 

 12in. behind this was reduced to -14, at 9in. -11, at 6in. -09, at 3in, 

 •07, and at lin. distance the square card received a pressure of only 

 •031b., or one-fifth of what it endured if the disc were removed. 

 The second was the apparent attraction of a large surface for a 

 small one behind it. This was found to be due to a current setting 

 towards the back of the surface exposed to the wind. Two ex- 

 periments illustraitng this i^eculiar effect may be mentioned. In the 

 first the 9in. disc was fixed in the current of air and a second disc 



