142 
HAYFOKD. 
Equatorial 
radius, a, 
in meters. 
Polar semi¬ 
axis, 6, 
in meters. 
Compres¬ 
sion 
( a-b )/ a. 
Bessel spheroid of 1841. 
6,377,397 
6,356,079 
1/299.2 
Clarke spheroid of 1866. 
Harkness, 1891. From ‘ ‘ Th e Solar 
Parallax and Related Constants,” 
Washington, 1891, p. 138. From 
6,378,206 
6,356,584 
1/295.0 
a variety of sources. 
The spheroid determined by the 
thirty-ninth parallel triangula¬ 
tion and the Lake Survey arc of 
6,377,972 
6,356,727 
1/300.2 
the meridian. 
The spheroid determined by the 
thirty-ninth parallel triangula¬ 
6,377,912 
6,356,309 
1/295.2 
tion and Peruvian arc. 
Eastern oblique arc of the United 
6,378,027 
6,356,819 
1/300.7 
States . 
Nantucket and Pamlico-Chesapeake 
arcs of meridian and Peruvian 
6,378,157 
6,357,210 
1/304.5 
arc of meridian. 
Lake Erie arc of parallel and Peru¬ 
6,378,054 
6,357,175 
1/305.5 
vian arc of meridian. 
6,379,822 
6,357,716 
1/288.6 
The last two determinations shown in the table are of 
light weight in comparison with the preceding three. 
A study of these values will show that the modem observa¬ 
tions in the United States indicate that the true value of the 
equatorial radius lies between the Clarke and Bessel values, 
but nearer the Clarke value, and for the polar semi-axis is a 
little greater than the^Clarke value. 
Having in mind the large number of astronomical stations 
attached to, and the large area covered by, the arcs already 
utilized in the United States, as indicated above, it is rea¬ 
sonably safe to predict that if the United States is eventu¬ 
ally completely covered by triangulation and astronomical 
stations are liberally supplied everywhere and the mean 
figure deduced from these observations alone, regardless of 
those made in other countries, neither the equatorial radius 
nor the polar semi-axis so computed will differ from the 
Clarke values of 1866 by as much as 500 meters, and it is 
about an even chance that either value will not differ from 
the corresponding Clarke value by more than 170 meters, 
