TRANSACTIONS OF SECTION A. 579 



2, On a Neiv Form of Instrument for Ohservinrj the Marjnetic Dip mid 

 Intensity on Board Ship at Sea. By Captain E. W. Ckeak, C.B., 

 F.R.S. See Reports, p. 29. 



3. Note on some Results obtained with the Self-recording Instruments 

 for the Antarctic Expedition. By Dr. R. T. Glazebrook, F.R.S. 



4. On a Determination by a Thermal Method of the Variation of the 

 Critical Velocity of Water -with Temperature. By H. T. Barnes, 

 M.A.Sc, D.Sc, Lecturer in Physics, and'E. G. Coker, M.A.^ D.Sc, 

 Assistant Professor of Civil Engineering, McGill Uniiiersity, 

 Montreal. 



The critical velocity, or point at whicli tlie flow of water tbrough a pipe 

 changes from stream-line to eddy motion, has been the subject of a series of 

 experiments by Osborne Reynolds from the philosophical as well as the practical 

 aspect. Two methods, which are too well known to require descriptioti, were 

 adopted in his experiments— the method of colour bands and the determination of 

 the law of resistance governing' the flow at velocities above and below the critical 

 velocity. From the results of his work Reynolds was able to verify certain mathe- 

 matical deductions as to the efl'ect of viscosity and diameter, which led to exceed- 

 ingly simple expressions for determining the change in the flow. The efl'ect of 

 temperatm-e was, however, less completely verified. In so far as the critical 

 velocity is dependent on the viscosity, the temperature coefficient of the viscosity 

 was taken as representing this temperature change. General experimental results 

 indicated, at least approximately, that the law of Poiseuille for the flow through 

 capillary tubes held for the critical velocity between 4° and 22° C. It was deemed 

 desirable by the authors, on account of the large effect produced by temperature, 

 to determine this coefficient directly by a new method, and more especially as the 

 law of Poiseuille itself was deduced from experiments ranging only as high as 



In the present paper a new thermal method of measurement is described, and 

 also experiments by this method with a brass pipe 0'414 inch in diameter at 

 different temperatures between 15° and 86° C, together with the general results 

 so far as it is yet possible to communicate them, showing the reformation under 

 perfectly steady and uniform conditions of the stream-liue flow at velocities very 

 much above the critical point measured by Reynolds. 



Thermal Method of Measuring Critical Velocity. 



If water be heated while flowing through tubes in stream-line motion, the distri- 

 Tjution of heat throughout the water column is not uniform. In the case where the 

 heat is applied at the outside of the tube, as in the experiments of L. Graetz, only 

 the few layers which are almost stationary in direct contact with the tube will be 

 heated, while the inflow water, which passes directly through the central portion 

 at a much greater velocity, will remain almost entirely unheated. In the case 

 where the heat is received from a central wire, the heat is carried off" by the 

 quickly moving water in a cloak as it were around the wire, leaving the sides of 

 the tube unheated. At and beyond the point where eddies make their appearance 

 in the flow, the entire column of water is mixed and stirred, and the temperature 

 distribution becomes uniform. The point of change, or the critical velocity, may 

 be then very clearly defined by observing the sudden increase in the temperature 

 of the floAving water. In some of the first experiments this change of tempera- 

 ture was observed by noting the increase in resistance of a platinum wire threaded 

 through the centre of the tube heated on the outside, and the preliminary results 

 showed that the presence of a wire of 6 mils' thickness in a tube of about \ inch 



