( 397 ) 



XV. — The Viscosity of Solutions. By C. Ranken, B.Sc, Carnegie Research Scholar, 

 and Dr W. W. Taylor. Part I. Communicated by Professor Crum Brown. 



(Received July 2, 1906. Read July 2, 1906. Issued separately December 31, 1906.) 



In a recent paper * published by the authors, some measurements of the viscosity of 

 aqueous solutions indicated that it would be of interest to investigate more fully the 

 viscosity of solutions, especially those which exhibit what is now generally known as 

 " negative viscosity," over a wider range of temperature. 



This has now been done to a certain extent, and to enable a comparison to be made 

 between electrolytes and non -electrolytes in aqueous solution, some other substances 

 have been included in the investigation. They are : — potassium chloride, potassium 

 chlorate, ammonium iodide, potassium ferricyanide, potassium ferrocyanide ; mercuric 

 chloride, mercuric cyanide ; carbamide. 



Method and Apparatus. 



In the determination of relative viscosity by Ostwald's method (Ostwald-Luther, 

 Phys. chem. Messungen, p. 260) there are two, if not three, points which require special 

 attention, in addition to the usual precautions to ensure constant temperatures, etc. 

 These special features are : — the cleaning of the capillary tube, the constancy of the volume 

 of liquid introduced into the viscosity tube, and the measurement of the time interval 

 which the liquid takes in flowing from the one mark to the other. The last is the most 

 difficult to deal with, and, in order to minimise the error, the dimensions of the tube 

 are often so adjusted by means of a very fine capillary, or of a bulb containing a com- 

 paratively large volume of liquid, that the time of flow is very long. Each of these 

 devices is open to serious objections, and, as our previous experience had convinced us 

 of the inherent unreliability of ordinary stop-watches, and especially of the impossibility 

 of starting or stopping them at the precise instant, we adopted a different method of 

 time measurement which has been found to be convenient and accurate to a high 

 degree. 



An electromagnetic clock with half-second pendulum electrically records seconds, 

 or other intervals of time up to one minute, on a band of smoked paper. This band 

 of paper is about 3 metres in length, and runs over two drums, one large and one 

 small, the large one being driven at an approximately constant speed by an electric 

 motor ; the speed is reduced and adjusted by means of a worm reducing-gear. An 

 electromagnetic time-marker records the experimental times on the paper immediately 



* Proc. Roy. Soc. Edin., 25, p. 231, 1904. 

 TRANS. ROY. SOC. EDIN., VOL. XLV. PART II. (NO. 15). 56 



