modulation of the reflected beam can be determined by adjusting
the variable electrical delay unit. The variable light delay unit
also shown in figure n is a method for calibration of the instru
mental zero errors and is only used occasionally.
By incorporating two crystals giving slightly different modulation
frequencies it is possible to avoid having to know the distance to
the far reflector accurately to start with. The procedure is very
similar in principle to the "method of excess fractions" used with
optical interferometry. Alternatively, if the distance from Geodi-
meter to reflector is very accurately known then the velocity
of light in the prevailing atmospheric conditions can be measured.
Of course, the refractive index of the air along the light path has
to be known if this velocity is to be reduced to a vacuum value, or if
a distance measurement is made based on a given vacuum velocity.
This refractive index is usually derived from measurements of
barometric pressure, air temperature and humidity, the values
being substituted in a formula based on laboratory measurements
(8).
In his latest paper Bergstrand gives a review of Geodimeter
measurements of the velocity of light and selects two values based
on weighted means of many determinations; these are, C0
299792.85 0.16 km/s and C0 299792.75 0.34 km/s; the former
set is derived from measurements with an improved form of
Geodimeter. The error limits quoted are smaller than the true
experimental variation because of the statistical procedure used.
Bergstrand states the precision of a single velocity determination
(itself the mean of several velocity measurements) to be about
T: 0.4 km/s.
I will now discuss one other microwave method for the precision
measurement of the velocity of electromagnetic waves which might
also be applicable for the measurement of distance. This is the
microwave interferometer (10), a device using interference of short
microwaves instead of optical waves.
Figure 12 illustrates the principle of the latest and most refined
form of this apparatus. The movable part of this interferometer
consists of a carriage about 7 m in length, supporting a pair of
receiving apertures ("horns") in line and facing outward towards
the two transmitting horns. Each transmitting horn is supplied
with microwave energy from a common source and is separated
from each receiving horn by an air space of about 10 m. The energy
received by each horn on the carriage is mixed with that from the
other to produce interference so that the detected sum of these
two signals undulates for each half-wave displacement of the car
riage. By making the carriage displacement through 970 minima
by means of an accurately known end-gauge a precise measurement
of microwave wavelength (ca. 4 mm) could be made. When this
wavelength is multiplied by the operating frequency of the instru
ment (72 Gc/s) and by the measured refractive index of the air the
io9
