OF ARTS AND SCIENCES. 



49 



Vi 



R^% 



In the calculation of the ratio - of this table, the value of ^ ^ used 



was the mean of that obtained from experiments 16 and 17, after cor- 

 recting for temperature. The agreement of these two values within 

 0.1 per cent is a test of the accuracy of the method, as the two ex- 

 periments were made on different days, and the direction of the current 

 was reversed. It will be seen that the vahie of this quantity increases 

 slightly with the temperature, as we should expect from the slight 

 ditFerence in size of the two tubes used. The values of x will be seen 

 to agree quite closely, with the exception of experiments 2 and 21. 

 I have treated these results in the same manner as those of Meyer, and 

 the result is shown on Fiij. 1- 



The point A is plotted from experiment 18. and i? from 21 ; so that 

 the lines AC and BC show the greatest variation in nine out of ten 

 determinations, while the majority of these lie so close together as not 

 to be capable of clear representation between A and B. The point C 

 has been raised from C for distinctness. Experiment 2 would indicate 

 a deviation from the straight line ; but I do not regard this as a per- 

 fectly reliable determination. More experiments are needed between 

 0° and 100° to establish the law. 



In order to compare these results with those of Meyer, I have been 

 obliged to assume his value of jy = 0.000168 at 0.° C. as a start- 

 ing-point, since the apparatus which I have used does not give absolute 

 values of the coefficient of viscosity, but only ratios. It would ap- 

 pear, however, that the great concordance among the results thus far 

 obtained would warrant its application to absolute measurements, for 



VOL. XII. (n. S. IV.) 4 



