24 SECTION PHYSICS. 



that there is perfectly distinct evidence in the Hydrographic Office 

 of the existence of this very remarkable result. 



Having said so much, and having pointed out that, in the theoretical 

 point of view, an exceedingly important correction was required in 

 the' Astronomer Royal's theory, I must also point out distinctly that 

 the Astronomer Royal himself was conscious of the fact that his 

 theory made an assumption which was not rigorously true. He 

 justified it by the consideration that the results did not produce effects 

 of considerable importance with respect to the practical problem. 

 But there is just one thing in which this way of putting the Astro- 

 nomer Royal's case must be accepted with considerable reserve. 

 That is, that if any attempt were made on his theory to calculate 

 beforehand what would be the effect of particular masses of iron, such 

 as bars or globes, the theory would be found to be greatly at fault. 

 But if we merely take the integral result of the whole iron of the ship, 

 then we get precisely the same effect when the iron is distributed 

 almost symmetrically, as it is in all ordinary ships, as if the particular 

 assumption referred to had not been made. So that unless we wish 

 to calculate beforehand the effect of particular masses such as bars 

 and spheres, we may accept the Astronomer Royal's conclusion with- 

 out reserve. In respect to the estimation of the nature of the 

 effects produced by particular masses of iron in the ship, Archibald 

 Smith's more complete theory has been very valuable. The Astro- 

 nomer Royal's theory for instance could not have accounted for the 

 great effects that may be due to magnetic induction upon a vertical 

 mass of iron such as the stern post of a ship or a vertical staunchion. 

 But setting that aside altogether, and leaving what no doubt the 

 Astronomer Royal was ready to admit was a subject for special 

 investigation, the effect of bars and so on, and the fact that they 

 differ enormously in virtue of the mutual action of the different 

 portions of the iron from what they would be were such mutual action 

 insensible, then I have no more to say, but that the Astronomer Royal 

 has given a complete and practical method for correcting the compass. 

 When we come to some of the modern ironclads, such as the 

 Inflexible, in which I believe there are turrets unsymmetrically 

 distributed, that is to say a forward turret on one side, say the star- 

 board, and an after turret on the port side, then the want of symmetry 



