TRANSACTIONS OF THE SECTIONS. 259 



Angles at which Eatio of forces at the tiller in 



rudder is held. Guinpel's and common rudder. 



0° -80 



5° -78 



10° -74 



15° -(56 



20° '56 



25o -44 



30° '.'.'.'.'.'.'.'.'.'.'..'.'.'.'.'....'.'. -29 



35° '13 



38° 40' -00 



The ratio of work required to put over the two rudders respectively to angle 

 88 deg. 40 min. is found to be four tenths if the resistance is supposed to vary as 

 the square of the line of inclination, and five tenths if (as is more correctly assumed) 

 it vary simply as the sine of this angle. 



The value of this rudder is manifest, as when the angle increases the relative 

 force dimiuishes considerably ; and after about 26 deg. the absolute force duuini.shes, 

 until at 38 deg. 40 min. it vanishes; and it is for the higher angles that the diffi- 

 culty of putting the common rudder over is most felt. 



APPENDIX. 



On the Effect, of Heat in altering the Molecular Structure of Steel. By W. F. 

 Barrett, F.R.S.E., Professor of Phi/sics, Royal College of Science for 

 Ireland. 



If a wire of steel of any thickness be heated, by any means, at a certain critical 

 temperature and for a certain length of time, the wire ceases to expand, although 

 heat be continuously poured in. During this period also the wire does not sen- 

 sibly increase in temperature. The length of time this abnormal condition lasts 

 varies with the thickness of the wire and the rapidity with which it can be heated 

 through its mass. 



Rods of steel from two to three tenths of an inch in diameter cease to expand for 

 five seconds when the wire is heated in a powerful combustion-furnace. This 

 change takes place as the wire begins to glow with a red heat and after it has ex- 

 panded one hundredth of its total length. The temperature of the critical point is 

 a little over the melting-point of antimony, but under that of silver. The heating 

 being continued, tlie elongation of tlie wire is resumed till the wire glows with the 

 full heat it can attain from the source ; it then ceases to expand, and no further 

 change is noticed till the heat is cut oft". Wlien this is done, the wire begins to 

 cool down regularly, till it has reached the critical point at which the change took 

 place on heating. Here a second and reverse change occurs. Tliis is the di.sturb- 

 anee first noticed some time ago by Mr. Gore, and wliich he believed was confined 

 to the cooling of steel wires and of wires of small diameter This, however, the 

 author has found not to be the case. Hence it appears that, at a certain critical 

 temperature, the molecular structure of steel changes. The specific heat of the 

 metfil doubtless increases at this point, and tlie heat that is rendered latent is again 

 given out at the con-esponding temperature during cooling, when both the molecular 

 structure and the specific heat of steel pass into their normal state at the tempera- 

 ture of the air. 



The molecular disturbances during the cooling of steel wires are thus exceedingly 

 remarkable, and may be summarized as follows : — 



1st. M, the critical temperature these wires undergo a momentary and consider- 

 able elongation, the amv>unt mainly depending on tlie strain to which the wire is 



