240 CAUSES OF DRIFT. 
with a high angle of elevation and a very low velocity what he calls the 
rolling effect comes into play, and he said that this rolling effect tends 
to carry the projectile to the right in the descending branch. Is that so? 
MAJOR-GENERAL OWEN—With a sufficient rotation. 
CAPTAIN MANSELL—I cannot think that that is so at all. I think 
that at those critical periods the tendency is to a slip of the projectile to 
the left. Some experiments of Professor Magnus with spherical pro- 
jectiles, which were published in the paper that the lecturer has referred 
to, completely bear out this point I think, and if I may | would just 
like to show what I mean. ‘ 
Fig. 1. 
ee A 
B \ 
( Lai eae SN cet 
us 
ae ee A 
He took a cylindrical body and rotated it, and directed a blast of air 
upon it in a direction at right angles to its axis of revolution. Let 
Fig. 1 represent the conditions. It is evident that the body in its rota- 
tion carries the air round with it in the direction of its rotation. Now 
we see from the figure that at B the current of rotation is opposed to 
that of translation (A), while at C the current of rotation is with that of 
translation. It follows therefore that the resultant velocity of the air is 
less on the side B than it is on the side C. Under these conditions 
Professor Magnus shewed by a beautiful experiment that the lateral 
pressure of the air on the body is greater on that side where the velocity 
is least, viz., at B in Fig.1. Magnus further shewed that when the 
velocity of the blast was great in comparison with the velocity of rota- 
tion the lateral pressure on the sides was not affected. This is easily 
understood ; for if V is the velocity of the blast and it is directed on 
the body at rest the ratio of the velocity of the air on either side is 
wal. Now if the body is rotated with a velocity v, the ratio of the 
aint, 
a, : 3 ; Vv : 
velocities of the air, on either side, becomes * and when V is very 
Wes 
great in comparison with v this ratio remains approximately units, and 
therefore the lateral pressures on either side of the body remain equal. 
This explains why some change from the gyroscopic cause of drift is 
not noticed except at critical velocities, such as are experienced when 
firing with low muzzle velocities at high angles of elevation. 
Iam unable to reproduce Magnus’ original experiment but if I may 
I will show you with a candle. If I blowa blast of air on the left hand 
side of this candle flame the velocity is greater on the left hand side 
