Gould's Report on Trans-Atlantic Longitude. 233 
sharp frost had thrown the otherwise defective line into a con- 
sufficiently numerous to ensure a trustworthy result; still this 
third link in the chain of longitudes is undoubtedly its weakest 
part. 
A further difficulty was presented by the unprecedented 
interval between the meridians; necessitating the use of simple 
clock-comparisons instead of star-signals, and preventing the 
interchange of observers for eliminating the effects of person 
equation. Star-signals—i. e., signals transmitted at the instant 
a given star passes the several wires of the transit instrument, * 
first at one station and then at the other, and registered at both 
—have these advantages over mere clock-signals—beats sent 
om each station and similarly registered—that they give 
results inrlepentens of the star’s right ascension, which, can- 
. 
the star's passage between the two meridians. Where this inter- 
val is large, as in the case before us, the special advantage o 
Star-signals mainly disappears; and even if it did not, they 
would require too protracted an oecupation of the cable, and in 
the climate encountered, or indeed, in any climate, the chances 
would be great, with an interval o ours, against the 
Same star being observed at both stations. - 
With the superior catalogue of time-stars, however, which 
had been prepared, and with the careful experiments for per- 
sonal equation which were made both before and after the expe- 
any, also, arising from uneliminated personal at 
this was true; in both cases, the discussion of these points in , 
the ‘eport seems conclusively to show : 
matic registration of the signals received ; inasmuch as the loss 
of time i: ine the si in the method employed, not 
athe Magee senels eee The most sensitive 
ission of signals with 
circuit during an ei 
deflection only was 
