152 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1911. 
The same general statement can be made relative to any of the 
standard forms of low-frequency telegraphy over wires as now 
practiced, such as the polar duplex, the differential duplex, and the 
duplex-diplex, employing alternating currents of low frequency and 
standard keys, relays, and sounders. 
Inserting a regular 150-ohm telegraph relay in series in the line 
cuts down the high-frequency current to a small percentage of its 
original value, which indicates the marked influence of the presence 
of iron in such a circuit. Furthermore, it was noted that at 100,000 
cycles the hysteresis of the iron core was so great that it became 
heated very perceptibly in a few moments. 
Since a portion of the telegraph lines now used is still composed 
of iron wires, it would be expected that electric waves would be 
propagated over such wires less efficiently than over copper wires, 
although it is well known that electric waves penetrate only about 
one-thirteenth as deeply into soft iron for a given frequency as into 
copper, but this is modified by the fact that the iron in telegraph 
wires is not soft iron and in addition is galvanized. 
[Section 4 of this paper, giving details of measurements of electric 
waves of frequencies from 20,000 to 100,000 cycles per second on a 
standard telephone cable line, is omitted from the present reprint 
by the Smithsonian Institution.] 
SUMMARY. 
Radiotelegraphy has no competitor as a means of transmitting 
intelligence between ships at sea and between ships and shore stations, 
and on land it-is also unique in its usefulness in reaching isolated 
districts and otherwise inaccessible points. To what extent it may 
be also developed to furnish practical intercommunication according 
to the high standard now enjoyed in thickly populated districts it 
is not attempted to predict. 
The foregoing experiments indicate that either the existing wire 
system, or additional wires for the purpose may be utilized for the 
efficient transmission of telephonic and telegraphic messages, and 
the former without interfering with the existing telephone traffic on 
these wires. 
The fact that each of the circuits created by the use of super- 
imposed high-frequency methods is both a telephone and a telegraph 
circuit interchangeably, makes it possible to offer to the public a new 
type of service, which it is believed will offer many advantages to 
the commercial world. This type of circuit should be particularly 
applicable to press association service, railroad service, and leased 
wire service of all kinds. 
The experiments described should not be interpreted as in any way 
indicating limitations to radio telegraphy and telephony in the future, 
for their present rapid development gives justification for great pros- 
