154 Velocity of the Galvanic Current in Telegraph Wires. 
graving tool, and partly to the bad quality of the paper of which 
the fillet is made. The Pittsburg register is in many plaees illeg- : 
ible owing to the shortness of the pass. This could be spared : 
with less inconvenience than any one of the others, but it is a 2 
very unfortunate circumstance that the St. Louis record, the most : 
important of all, should be so indistinct. It is affected, moreover, 
with the additional disadvantage, that the motion of the fillet 
was slow, and the liability to error in measuring thus greatly in- 
creased. 
The labor of measuring off the intervals on the fillet is great 
and very tedious, and the average error of reading amounts to 
some hundredths of a second, although the intervals were mostly 
measured with dividers and a metallic scale, not with a diagonal *, 
scale of horn. ; - 
“All the results adduced depend upon the mean of a great 
number of readings, more than 5000 measurements of the work 
of February 4 having been made, in order to obtain them. The a; 
ging and zigzag course of the wires, but may be used in the ab- 
sence of positive knowledge on the subject, and gives the follow- 
ing distances for the several stations as measured on the wire. 
Washington, 
288 _— Pittsburg, 
622 334 Cincinnati, 
447 45 13 Louisville, 
1045 457 423 289 St. Louis. 
The experiments occupied a period of several hours, the Sea- 
ton clock graduating the scale, at Seaton-station, Washington 
City, Pittsburg, Louisville, and St. Louis. The operator at St. 
Louis gave arbitrary signals by breaking cirenit from time to time, 
at intervals of two or three seconds, which signals were recorde 
at all the stations. This he did for two successive minutes, 
twice during the evening. The same was done by the operat 
at Louisville, Cincinnati, and Pittsburg. Signals were also 
at the latter station for ten minutes after the Seaton battery wa 
+ Proc, Amer, Assoc., 1849, p. 189. 
