HOGG. — FRICTION AND FORCE DUE TO TRANSPIRATION. 135 



in which the fibre is. Great care is taken in setting to turn the 

 magnet in one direction so that any movement which the magnet, and 

 therefore the pointer, may make, which is not made by the armature, 

 may be eliminated. It was found, however, that with an ordinary 

 permanent magnet this elimination was not satisfactory, so that a 

 strong electro-magnet was finally used. This proved quite satisfactory, 

 and certainly the error here is not so great as that arising from the 

 innate difficulty of setting the vane to the proper position. To deter- 

 mine when the vane is in the proper plane a telescope is focussed on 

 a scale reflected in the mirror, E, and the point on the scale which 

 corresponds to the proper position of the vane is noted. This, of 

 course, remains the same so long as the disposition of the apparatus 

 is unchanged. 



From what has been said previously on the relation between the 

 force on the vane and the size of the transpiration space, it follows 

 that for any given pressure the force on the vane in this instrument 

 will be greatest when the vane and annulus are in the same plane, for 

 then the annular space is least. The object is to measure the force 

 when this disposition is secured. Since the blackened surface is 

 turned towards the heat source, when this surface is irradiated the 

 vane recedes, because pressure grows on the hot side. By turning the 

 magnet, enough torsion is given to the fibre to bring the vane forward 

 again into its zero position. If too much torsion is given, and the vane 

 is thrown in the least past the annulus, then it continues to swing 

 forward, because the force on it diminishes as its distance from the 

 annulus increases. A glance through the telescope shows when this 

 has happened, and in this case the magnet must be turned back and 

 another trial be made. Much practice is required that one may set the 

 vane, even with moderate success. The difference between the read- 

 ings on the circular scale when the vane is and when it is not irradiated 

 is the torsion necessary. As is allowable, the force on the vane is 

 assumed to be proportional to the angle of torsion when the latter 

 is not large. 



In this maximum effect we have a method of finding the zero posi- 

 tion without actually looking to see that the vane and annulus are 

 co-planar. The method consists in illuminating the vane, setting to 

 the maximum point, and noting the mark indicated by the cross-hair 

 of the telescope. This mark indicates the zero position. It is really 

 best to determine the two balancing points for any angle of torsion, 

 one when the vane is behind the annulus and the other when it is in 

 front. As the torsion on the fibre is increased the places of balance, 

 indicated by the cross-hair of the telescope, approach each other, 



