1904.] 



on the Motion of Viscous Substances. 



489 



^It will be seen that just at first it turns more rapidly than after- 

 wards when it settles down to a uniform mte of twisting. It is 

 apparently possible to keep on turning the rod as long as one pleases. 

 By painting a white line down a rod, and then twisting it, a beautiful 

 spiral line is produced. On removing the twisting force the rod turns 

 back a short distance, at first rapidly, but slows down gradually 

 to rest. 



Traction Method. 



A very interesting question arises as to how the material moves in a 

 rod of pitch or such-like substance when drawn out. If we subject 

 a rod of pitch to traction, say by suspending it from one end and 

 hanging weights from the other end, we find that it draws out at an 

 approximate uniform rate. The rate is at first a little faster than it 

 is later on ; it however finally settles down to a 

 uniform rate, provided the tension is kept the 

 same. This is similar to the effect observed in 

 the case of torsion. Also, as in the case of torsion, 

 we get a slight recovery on removal of the force. 



The exact way in which the particles move in 

 a rod as it is drawn out is not at all clear. In 

 the case of the flow through a tube, we know 

 that the centre flows faster than the outer parts 

 because the sides are held back. Nothing in the 

 case of the rod corresponds to this. To try and 

 observe the character of the flow in a rod the 

 plan was tried of drawing out a rod made up of 

 two shorter rods of different colours, but otherwise 

 alike, joined end to end. The junction was made 

 as sharp as possible and lay at right angles to the 

 axis of the rod. Rods of shoemaker's wax and of 

 glass were used. Difficulty was found in getting 

 two differently coloured glasses of exactly the 

 same fusibility, and also in forming a really sharp 

 line of junction between them. As far as the 

 observations go, they show at least that the particles lying in a plane 

 do not move so as to lie on a curved surface, such as occurs in the 

 flow through a tube. 



Fig. 5 is intended to show how a small quantity of the material, 

 included between two near planes, lies after a short time. A small 

 cube of the material is shown in the initial stage and its subsequent 

 shape when drawn out. The connection will thus be appreciated 

 between the kind of movement here taking place, and that consistent 

 with the definition of viscosity as illustrated by the model. 



The rate of elongation of the rod divided by the tension gives us 

 a coefficient of viscous traction. We have every reason to expect 

 this to be about three times the viscosity of the material, and ex- 

 periment supports this view. The value of the viscosity of pitch 



I. 



I 

 F ] 



I 





Fig. 5. 



