advantage of the Strasser/Schwendener method is the compara
tively simple observation and reduction procedures recommended.
Initially, the field work was envisaged as likely to take two
weeks, but as the tests progressed, unexpected behaviour anomalies
of the gyro-attachment were found to occur at surprisingly regular
intervals. This new development necessitated a complete reappraisal
of the investigation and eventually resulted in over two-and-a-hail
months field work being carried out in order to produce and verity
the results and conclusions presented in this paper.
At the commencement of the tests, a number of daily run-ups
of the instrument were made, during each of which approximately
eight turning points were observed for purposes of azimuth deter
mination. The then conventional method of deriving the azimuth
from the Schuler mean [5l was applied and although several
groups of data revealed remarkably consistent sets of observations,
a few had isolated fringe observations which were apparently
inconsistent with the body of data recorded. A typical example is
shown here.
In this example, one might, at first glance, be inclined to discount
either, or both, the first and sixth Schuler means because of the
apparent consistency existing between the second, third, fourth and
fifth means. In fact, the sixth Schuler mean gives the best value
of gyro-indicated north.
As a result of the above, the authors decided to prolong the
observations of the gyro-oscillations beyond the eight turning points
and to determine what trend, if any, developed over a longer period
The number of turning points was increased from eight to sixteen,
from sixteen to thirty, and eventually it became common practice
to observe forty-five turning points during a single run. Since this
involved two-and-a-half hours of extreme concentration (an
excessive task for one observer) it became necessary to switch
observers after every twelve to fifteen turning points.
272
Left
Right
Mean
i79°i6T9"
i8o°09'5o"
179.16.37
i8i°o3'o2"
179.16.55
181.02.10
180.09.33
179.17.51
181.01.17
180.09.34
179.18.47
181.00.31
180.09.39
179.19.33
180.59.45
180.09.39
179.20.19
180.59.28
180.09.53
180.59.11
The numbers in italics are the means of two successive left or rlght
observations; the mean of such a mean and the associated observation is the
Schuler mean, printed in the third column.
