1874.| The Iowa and Illinois Tornado. 385 
petal,—to wit, the spiral. The illustrations give copious 
ocular demonstration that such was the case. The ruins of 
houses, whenever they could be distinctly traced, proceeded 
in the curve towards the centre, which apparently was always 
reached before the completion of an entire revolution. This 
does not imply that the centripetal winds never performed a 
complete revolution before reaching the centre, but only that 
from within about too yards of the centre this did not take 
place. Ruins carried from that distance seemed generally 
to reach the centre in about two-thirds of a revolution. 
This led me to assume at about 100 yards from the centre 
the wind generally made an angle of about 25° with the 
tangent. This gives a centripetal component compared 
with the circular as 0°42262 is to 090631. 
Besides its horizontal motion, the wind had necessarily an 
ascending motion, which reached its maximum at the centre 
and its minimum at the circumference. The necessity for 
this lies in the proof of the centripetal motion of the wind. 
Otherwise the centripetal motion of the wind would require 
to double itself at half the radius from any point toward 
the centre : this would soon give it a motion infinitely great, 
which is absurd; there would also be no means of escape 
for it at the centre. To a great many excellent witnesses 
the funnel presented the appearance of a screw with its 
thread running upwards. In addition, all objects within the 
sphere of the tornado seem to have been partly lifted, partly 
pushed forward. We have seen that at A. Gibson’s house 
the arms made an angle with the perpendicular of probably 
about 45. Since these arms, as there is every reason to 
believe, were inclined in the normal direction of the winds, 
which is that of least resistance, we are justified in assuming 
that in the black funnel itself the inclination of the winds 
to the horizon was at least as great, and probably greater. 
The Velocity of the Wind.—This is a very difficult topic ; 
the task would have been much easier if we had been 
dealing with a horizontal wind merely. If even the perpen- 
dicular and horizontal components were invariable, the hope 
of a tolerably accurate solution of the problem would be 
much greater; but while the former decreases from the 
centre to the circumference, the latter increases. In order 
that no chance of a solution might be lost, I give the fol- 
lowing from data I collected :— 
1. A school-house, weighing 30,000 Ibs., was pushed 
25 feet from its foundations. The surface exposed to 
the wind = 360 square feet + the roof, which hada 
slant of 45°. 
