266 
PRESTON. 
to the regular work of every trigonometrical survey of a 
country, I shall call attention to a subject already referred 
to, which promises to furnish novel results in the near future. 
Y. The Variation of Latitude. 
The International Geodetic Association now proposes to 
establish four stations as near as practicable at equal dis¬ 
tance around the earth and all within half a mile of the 
& same parallel of latitude. When it was discovered that the 
latitude changed, and that we were at one season of the year 
about 60 feet nearer the equator than at another, the result 
was so startling that many observatories engaged in the 
work of observing the latitude. The outcome was that the 
motion of the earth’s axis could be traced with reasonable 
accuracy from about 1888 to the present time. But the 
trouble is that the result was to a certain extent vitiated by 
the fact that the star places were uncertain, and although by 
an ingenious method of combining the observations this de¬ 
fect to a large extent disappeared, nevertheless the observa¬ 
tions did not yield the desired precision. Now it is proposed 
to so choose the stations that the results will be entirely un¬ 
influenced by any errors in the accepted position or proper 
motions of the stars. This can be done only by locating all 
the points of observations on the same parallel of latitude. 
Then the further question arises, at what distances apart * 
shall they be? Mathematical considerations show that the 
most accurate determination of the quantities in question is 
attained when the stations are equally distant around the 
earth. This would seem to be the natural supposition. 
Mathematics, after all, is only common sense. So that if 
three stations are adopted they should be 120° apart; if four, 
90°. But there comes in the condition that the configuration 
of the continents will not permit an equal distribution, so that 
a compromise and adjustment has to be made between the 
mathematical, physical, and social conditions. The impor¬ 
tance of the last named must not be underestimated. We 
may have trained observers, but they cannot make the best 
