THE DRIFT OF SERVICE PROJECTILES. 
475 
effect of rolling* a ball on a cushion of air, the roll would be to the left instead of 
to the right; but I have noticed with a tennis ball that when you are serving 
with cut the-ball goes pretty straight at first; it is only when the velocity is lost 
that it takes any drift at all. Whether from that you can argue that with a high 
velocity there is little or no friction, and therefore little or no either roll or slip I 
do not know, but it seems rather to point that way ; it seems true that the velocity 
has to drop before the effect can be seen, before drift sets in in the case of a golf 
ball or a tennis ball, but that, as I say, I cannot speak about for certain. 
Captain Mansell talks about rapidity of spin and rapidity of precession. I did 
not intend to make allusion to it at all. All I say is that whatever the rate of 
precession is the shot has got to precess faster than the rate at which the 
trajectory curves in order to get down at all. If it does not precess as fast then 
the spin will eventually lose control over the point and it will fall further and 
further back and finally turn broadside and, so far as I can make out, it is only 
when the shell turns broadside that you get this tremendous drift to the left. I 
think, Captain Osborn, when the shell is at a high, angle and drifts to the 
enormous amount of 1300 feet to the left it must have been after it turned broad¬ 
side and not before. Until the shot turns broadside it will have a right-handed 
drift. Of course when I am talking of not seeing anything less than an angle of 
10°, I mean simply when the shot is in flight one is, of course, some way off it, 
and seeing it at a considerable distance. I daresay if you took the perforation of 
shot through any substance you might form a closer estimate. 
As regards flat-headed shot J do not like to touch upon the subject. In the 
pointed shot I take my stand upon the fact of the resistance of the air forcing the 
point away. With flat-headed shot I do not think it is quite decided whether the 
resistance of the air will force the head of it away or force it back again when it 
is slightly displaced from tjhe liue of motion; but if, as General Owen has said, 
with this machine it has been observed that the point of flat-headed shot is to the 
left and the point of pointed head shot is to the right, if the flat-head has its 
point to the left in flight then there should be left-handed drift, because that 
means that the precession is left-handed. 
With regard to what Captain Mansell said about a straight line trajectory, 
that was only to simplify the theory for the purposes of explanation. One does 
not get it unless one does fire either straight up vertically or straight down 
vertically. That was merely to simplify the explanation at first, to accentuate the 
difference, to show what would happen if the trajectory was either very straight 
or very curved. The idea of the shot not completing one complete revolution 
in its trajectory comes from observing low velocity shell at high elevations, 
and because it has never been shown that if a revolution were completed 
that the preponderance would be to the right. This I have proved and, 
moreover, if the precession do not complete a revolution, there is no ex¬ 
planation of the periodic sound in flight. In low velocity howitzers, on the 
other hand, a revolution may not be completed, as I have shown above. As 
regards the quantitative aspect of the question, a mathematical analysis is, I 
believe, out of the question. An approximation may be arrived at possibly on 
the lines given us by Professor Boys, but what I have endeavoured to show in 
this paper is that the point is chiefly on the right and the reason of this, also that 
the variations due to high and low velocity, etc., vary with the amount of drift 
and that, therefore, the effect of the point being to the right chiefly must be a large 
factor in the determination of drift, if it be not the sole cause. 
Note .—Captain Mansell states, that the greater the velocity of rotation due to 
rifling, the slower will the precessional movement be, and that, therefore, accord¬ 
ing to my paper, the drift ought to be greater. This is perfectly true, and is 
