TRANSACTIONS OF SECTION A. 581 



to a size of pipe equal to theirs, and the other showing that the observations of 

 Reynolds between 4° and 22^ C. give a closer agreement with the authors' temper- 

 ature formula than with the formula of Poiseuille. 



The law showing the dependence of the critical velocity on the temperature 

 obtained by the authors may be stated thus : — 



p =/ (T) = (1 + -osooT + -oooroi-T) - ' 



between 15° and 86^ C. ; while the law of Poiseuille reads : — 



(1 + -033681 + -002211^)-' 

 between 0° and 45° 0. 



Experiments on Stream-line Floiu at High Velocities. 



It was found further that the unusually steady conditions obtained in the large 

 tank conduced to some interesting results in regard to stream-line flow at high 

 velocities. For certain sizes of pipes, over half an inch to as large as the authors 

 have yet used, i.e., 2^ inches, the flow re-formed again to stream-line above the 

 critical point of Reynolds, and persisted apparently as the stable flow to velocities 

 ranging from 12 to 20 feet per second. Beyond these velocities they were unable 

 to go, but in some instances no sign of breaking down occurred at these points. 



Two experiments were tried, which illustrate clearly that water flowing with 

 a perfectly smooth, unruffled surface is in stream-line motion. A circular orifice 

 was inserted in the side of the tank, which gave a clear rod-like jet of water that 

 issued horizontally under a high head and curved in a parabolic arc under gravity. 

 After all initial disturbances had died out in the tank a colour band was intro- 

 duced by bringing the colour tube to within about 3 inches of the centre of 

 the orifice. A clearly defined and sharp line of colour threaded its way through 

 the jet of water, shifting slowly from centre to side and back to centre again, 

 affected probably by slight movements in the tank. This thread of colour was 

 distinctly visible down to the point where the jet of water impinged against th« 

 waste weir, a distance of 15 feet. By introducing an excess of colour a similar 

 phenomenon to the breaking down of the stream-line flow in a tube was noted, 

 and the jet became suff'used with colour, broken, and unsteady up to within a foot 

 or two of the orifice. On reducing the quantity of colour the stream-lines 

 re-formed and the water became smooth, clear, and steady, threaded by the sharp 

 line of colour as before. Two sharp-edged orifices were tried, '^ and 2^ inches 

 diameter respectively, with coefficients of discharge equal to O-yTO. With the 

 heads used the highest velocity reached by the outflowing water, calculated in 

 the usual way from the formula 



V = 0-970 v/2^ 



was 30 feet per second. 



5. The Interference and Polarisation of Electric Waves. 

 By Professor G. Quincke. — See Reports, p. 39. 



6. On the Effects of Mapietisation on the Electrical Condaciiviti/ of Iron. 

 and Nickel, By Guy Barlow, B.Sc. 



_ The object of the experiments was to determine whether any simple relation 

 exists between the change of electrical resistance and the intensity of magnetisation 

 in iron and nickel wire when magnetised longitudinally. The effects of hysteresis 

 as shown by the magnetic change of resistance were also examined. 



The Wheatstone Bridge method was employed, with a bridge wire of low 

 resistance. The experimental wire was wound longitudinally on a thin rod of 



1901. Q Q 



