1862.] 27 



appeared to indicate that such was probably the case, by demonstra- 

 ting that not only were the vertical deviations from such models the 

 reverse of those in large guns, while the lateral deviations were the 

 same, but that it followed from this there must be a length and calibre 

 from which, while the lateral deviations still remained constant, the 

 range would be the same whether the centre of gravity of an eccen- 

 tric sphere were put " above " or " below." What, then, becomes of 

 the theory that the lateral and vertical deviations are due to the same 

 proximate cause, i. e. eccentric rotation through the air, and that it 

 is by the air, as assumed by M. Magnus, that both these deviations 

 occur ? The later disk experiments, t. e. from the gun, show that in 

 three out of the four positions of the centre of gravity in one plane (a 

 vertical plane) there is no decided rotation in such plane, or indeed 

 regular rotation in any plane ; yet these three positions in spheres all 

 give different ranges. 



Do these spheres rotate ? or can they in such case rotate in a ver- 

 tical plane with velocity enough to cause a vertical deviation, assum- 

 ing that the mean length of a range admits of being increased in this 

 manner, in opposition to gravitation? 



Perhaps the approximate causes may be looked for (of the two 

 kinds of deviation), the one more within, and the other chiefly out- 

 side the gun. 



As regards the longer ranges due to centre of gravity "above," may 

 not such increase be due to the fact of the nearer coincidence of that 

 important point with the axis of the bore ? In fact, ma^ not the ver- 

 tical variations in range be chiefly due to causes of a more directly 

 dynamical nature than has been generally thought ? while, respect- 

 ing the lateral, M. Magnus's views, founded on his experiments with 

 rotating bodies, appear not only incontrovertible of themselves, but 

 the identity of such deviations in models with those of large guns 

 offers no new fact on which exception could be taken or any new 

 question raised, which cannot be said of longitudinal deviations. 



To find whether a disk prevented from rotating in the bore, but 

 still delivered at the muzzle with centre of gravity "above," 

 would rotate in, and in such case by, the air, I made two disks (with 

 a straight edge above and below), the disks being very eccentric by a 

 transverse hole through the lower part. 



They both struck irregularly in any position, as seen by the 

 wood of the target, which shows also the form of the disks. 



