160 Velocity of the Galvanic Current in Telegraph Wires. 
All the results of the experiments made by the Coast Survey 
for the determination of the velocity of the galvanic current, have 
been most kindly placed at my disposal by my friend Mr. Walker 
to whom you had confided the entire direction of the work. 
These furnish the materials for a series of tables, containing the 
deductions from all the experiments. Collecting in one table all 
the cases where the current must have passed through the wires, 
and in another all those instances where the ground furnished 
part of the shortest circuit, we ae two equations with two un- 
known quantities, viz., the velocity in the ground and in the 
wires. Determining the latter tiene we find the velo- 
city deduced from all those cases where the distance is shorter 
through the wire, to be, by twenty-six different comparisons de 
pendent on 768 ‘readings, 15600 miles a second. Substituting 
this value in the other equation, we obtain the number of miles 
of wire to which the time of transmission corresponded. ‘These 
equations may be formed by aid of the method of least squares. 
The discussion of the results of all the experiments of the Coast 
Survey, namely, those on the lines from Washington to Cam 
bridge, Washington to Cincinnati, Washipgton to Charleston, Hp 
Boston to New York, show that the pes ee of a transmission 
through the ground, with a constant velocity, does not materially 
improve the accordance of the observations. Indeed, in the ex- 
periments of February 4, the only indication of a transmission 
through the ground is to be found in the fact, that the velocity 
derived from the St. Louis observations can be made to accord 
better with the mean of the other values, by assuming that the 
signals traveled 1030 miles instead of 1045. ‘The probable error 
of our estimate of the length of the wire is much greater than 
is. 
The cases depending on 920 measurements, in which the 
shortest route is through the ground, are twenty-two in number. 
The velocity deduced, on the assumption that the signals tra- 
versed the ground, would be 11,200 miles in a secon 
From all these considerations I infer that in the St. ‘Louis and 
Washington experiments which were, of all that have been made, 
the most favorable for exhibiting the phenomena,—the signals 
were in no case transmitted through the ground. 
We have thus endeavored to take account, as far as possible, 
of all the sources of error which cannot be avoided, and to escape 
all those which can. Our results, obtained from ‘different ie 
than 20,000 or less than 12,000 miles per sec 
From the combination of all the Coast Saree experiments 
with the electro-magnetic telegraph I have endeavored to 
a measure of the velocity which shall be as reliable as the 
