ON NA VAL ARCHITECTURE. 303 



which serves as a tow rope, and the extension of the spring balance 

 will, in such a case, correctly represent the resistance experienced by 

 the model at whatever speed it may be moving for the moment, which 

 must be exactly equalled by the whole towing force which is being 

 administered to the model. The extension of the spring is recorded, 

 on an enlarged scale, on the revolving sheet of paper, that is to say, 

 an arm connected with and representing on a large scale the extensions 

 of the spring, makes a trace on the paper as it revolves, and, at the same 

 time, a piece of clockwork connected with an electric circuit makes 

 a contact, which inscribes a mark on the cylinder at each consecutive 

 half second, and by the intervals between these marks, measured by 

 the distance scale, we see how much distance is moved in each half 

 second, that is to cay, the speed at which the operation is conducted j 

 we thus obtain a graphic record of the exact resistance experienced by 

 the model and of the speed at which she is moving. I should mention 

 that it is necessary, besides pulling the model by this spring balance, 

 to guide her very inexorably, so that she may not swerve from her 

 course. Because it is a curious circumstance connected with" the 

 theory of the motion of cuch a body through the water, that if you start 

 a perfectly symmetrical model or ship freely on a rectilinear course, 

 though it will perhaps run straight for a little distance, it most likely will 

 suddenly begin to swerve, and will then very quickly turn broadside 

 on. It is not sufficient to lead the head end inexorably in a straight line. 

 If you tow a sixteen foot model by her nose inexorably in a straight line, 

 you would, probably, expect that the stern would follow in the straight 

 line, as a flag will follow the mast. But instead of that being the case, 

 after the model has gone a little distance you will see its stern swerve to 

 one side or the other indifferently, and there remain at an angle of 

 5 Q , 6, or even 10 out of the straight line, acccording to the form, 

 with a curious set of eddies on the two sides of it. To resist this 

 tendency, the only constraint to which the model is subject, is applied 

 at each end by a delicate knee-jointed nicely counterbalanced frame, 

 which allows its head and stern to rise or fall and move backwards or 

 forwards with perfect freedom, so that it does not experience the 

 slightest pressure, except that which restrains its desire to swerve 

 from its course, and this force being transverse and not in the line of 

 motion does not affect the model's resistance. 



