6 
HAYFORD. 
of operations is that the actual motion of the pole shall be 
determined by simnltaneons observations at four stations 
widely separated in longitude, at each of which an instru¬ 
ment of the highest precision is to be so used as to guard, as 
fully as possible, against all known sources of systematic 
error. The details of the plan have been worked out very 
carefully, and it is admirable in every respect. 
The four stations selected are Gaithersburg, in Maryland, 
in longitude + 77° ; Ukiah, California, in longitude + 123° ; 
Mizusawa, in Japan, in longitude + 219°, and Carloforte, on 
the small island of San Pietro, just west of Sardinia, in 
longitude -f- 351°. One of these stations occurs in each of 
the four quadrants of longitude, reckoned from Greenwich. 
All are within three seconds of the parallel of 39° 8' 10". 
The conditions considered in selecting them, other than 
those indicated above, were that there should be nearly sym¬ 
metrical conditions as to the character of the surface north¬ 
ward and southward of the station to avoid unsymmetrical 
refraction; that the hygienic, social, and climatic conditions 
should be such that the observer might remain healthy, 
comfortable, and contented; and finally in Japan and at 
the European station the region was carefully studied with 
reference to the probable frequency of earthquakes. 
Two extra volunteer stations have now been added to 
these four, one at the Cincinnati observatory, which happens 
to be nearly on the parallel of the four selected stations, 
and one at Tschardjiu, in Turkestan. 
The four principal observatories have been constructed 
with the utmost care, under specifications furnished by the 
International Geodetic Association, specifications designed 
to insure that observations shall not be vitiated by local dif¬ 
ferences of temperature. The observatory proper is sur¬ 
rounded by lattice-work to protect it from direct sunlight 
and the roof is double. During observations the roof is 
rolled back to leave an observing slit two meters wide. The 
walls are so low that the telescope projects above the roof 
during observations. This fact, together with the unusual 
