Electrical Signals on Land Lines. 



417 



Tables II. and III. contain the results of two more sets of 

 observations on a line 550 kilometres long. M. Guillemin's 

 paper contains many other Tables ; but those selected appear fair 

 examples. 



Table II. 



Times, in ten-thousandths "1 

 of a second J 



Deviations, in degrees , 



Percentage of maximum 

 deviation 



Times expressed in func- 

 tion of a according to 

 the theoretical curve ... 



Value of a obtained by di- 

 viding the numbers in 

 the first line by those in I 

 the fourth J 



17 



33 



50 



65 



103 



136 



146 



162 



170 







2 



12 



29 



42 



43-7 



4: 



45-5 



46 







4-2 



25 



61 



88 



92 



93 



96 



97 



1 



1-8 



3-2 



5-7 



100 



110 



115 



130 



140 



17 



18-3 



15-6 



11-4 



10-3 



12-4 



12-7 



12 



12 



In Table II. it may be suspected that the effect of imperfect 

 proportionality between the higher galvanometer-deflections and 

 the current makes the agreement between the theoretical and 

 experimental curves better than it should be. This impression 

 is confirmed by the result of Table III., expressing the results 

 of experiments in which the arrangements differed from those 

 which gave Table II. only in the resistance of the circuit of deri- 

 vation through the galvanometer — a change which diminished the 

 range of the deflections, but can have produced very little altera- 

 tion in the true received current. 



Table III. 



Times, in ten-thousandths "1 

 of a second J 



Deviations, in degrees 



Percentage of maximum 1 

 deviation j 



Times expressed in func- ] 

 tion of a according to I 

 the theoretical curve ... J 



Value of a obtained by di- 1 

 viding the numbers in 1 

 the first line by those in j 

 the fourth J 



17 

 0-5 

 2 



1-5 

 11-4 



33 



1 

 4 



1-8 

 18-3 



50 



4 



16 



2-6 

 19-2 



65 

 10 



40 



41 

 15 8 



103 



20 

 80 



80 



12-8 



136 

 23 



92 



11 

 12-3 



150 

 24-5 



98 





This series, like Table I., would give a curve considerably 

 flatter than that due to the simple theory. 



The effect of imperfect insulation would not bring the theore- 

 tical curves nearer toM.Guillemm/s results. Imperfect insulation, 

 instead of flattening, rather sharpens the curve, as may be seen 

 from the following Table, calculated by the series given above 



