ATMOSPHERIC FRICTION. 
249 
aft, as shown in figure 1, to protect it from end-thrust. The 
plane is 16 feet long, 4 feet wide, and 4 inches thick. The 
shields are made of sheet zinc, their cross-section measur¬ 
ing 4J by 12 inches inside, and each shield closely envelops 
one end of the plane, yet has ample space farther within 
to allow the air to flow very freely from one shield to the 
other, through a large connecting pipe underneath the floor 
of the tunnel, thus equalizing the pressure. This pipe, or 
flue, measures 1 square foot in cross-section. 
The static pressures in the two wind-shields deserve careful 
attention. If they are equal, the resultant end-thrust is 
nothing, and the only deflecting force on the plane is the 
friction of the air along its sides. But in practice there is a 
difference of static pressure, which is measured by connect¬ 
ing the shields, by means of rubber hose, to a differential 
pressure-gauge graduated to millionths of an atmosphere, 
and usually read to one-ten-millionth.* Computing the end- 
thrust from the differential pressure, and adding or subtract¬ 
ing the result, gives the total skin-friction on the plane. 
The correction thus introduced is about 5 per cent, of the 
whole deflecting force. 
Considerable care was taken in the design of the plane to 
make it light and keep it perfectly straight. A frame was 
made of organ tubes and covered with paper in such a way 
as to be adjustable for warpage. As shown in figure 2, the 
paper is glued, not directly to the organ-tube frame, but to 
J-inch boards which slide over the four outer faces of the 
frame. As the paper was fastened on wet, it now remains 
very taut on all but the dampest days, and of course holds 
the sliding pieces firmly to the frame. The process of ad¬ 
justment is as follows: The two end sliding pieces are set 
vertical by means of plumb-lines, thus bringing the four 
corners accurately into a mathematical plane. The four 
corners are then joined by tight threads and the other sliding 
pieces tapped into line with a mallet. The operation re- 
*See “ Measurement of Air Velocity and Pressure,” Physical Review, 
December, 1903. 
