48 ATTEND IX. 



" The apparent patli of the sun in these high latitudes is but slightly inclined to the horizon ; and the 

 azimuth of any object was determined from the transit of the sun's first and second limb over the vertical 

 wires of the instrument. The time being known, the azimuth of the zero of the limb is easily calculated, 

 and nothing remained but to measure the horizontal angle between that direction and any object the 

 astronomical bearing of which was desired. The azimuth is reckoned from north by east round to 300°. 

 As objects for azimuthal determination, well defined glaciers, bluffs, islands, prominent capes, and the 

 most distant headlands were selected ; and, in order to make sure of the stability of the instrument 

 during the jieriod of observation, a second set of observations of the sun for azimuth of zero of limb 

 was obtained. 



"By means of two positions thus determined, a number of objects were located by the intersections of 

 the bearings of the known points, and whenever practicable a third or check azimuth was obtained ; in 

 this latter case any discrejjancy was properly taken into account according to known principles. 



" In observing with the sextant for altitude of the sun, the usual precautions were taken, and in par- 

 ticular the parallelism of the upper and lower surfaces of the covering-glass of the artificial mercurial 

 horizon was tested. An error of ten seconds, it is thought, cannot exist on this account, although 

 another roof gave results differing as much as fifteen minutes (?) in the direct and reversed position, and 

 consequently had to be rejected. 



" The sextants used were made Ijy Gambey, and divided to ten seconds. They were provided with an 

 astronomical telescope, which has invariably been made use of in connection with the artificial horizon. 

 When observing for latitude, multiplied observations were generally taken : first, three of the sun's upper 

 limb ; next, three of the lower ; and finally, again three of the upper limb. These observations were 

 commenced eight or ten minutes before noon. The corresponding index error was always determined. 



" Observations for longitude were never made nearer than three hours from noon ; and, whenever 

 weather and time permitted, corresponding observations in the forenoon and afternoon were secured. 

 On these occasions twelve observations, divided into four groups, and an equal number for the ui)per and 

 lower limb, wore taken. In observing corresponding altitudes, the index was set to an even five or ten 

 minutes, and the time noted when the contact was perfect. The successive changes of the index were 

 regulated according to the sun's relative changes in altitude. * * * 



* * * * * " In working up the observations, index error, refraction, and change of 

 the sun's declination, during the interval, were properly taken into account. 



" In a few instances, when the weather or other causes prevented an observation for latitude at noon, 

 two sets of oljservations were taken, as far distant from oue another as practicable, and latitude and 

 longitude deduced accordingly. Such was the case at Fiskernaes and Refuge Inlet. This method proved 

 very accurate, provided one set was not more than two hours from noon, and the other at least two 

 hours distant from the first. 



" Time was noted by a pocket-chronometer, which was compared before and after each set of observa- 

 tions with four box-chronometers, the rates of which had been determined at Xew York before leaving 

 port. At St. John's, Ne^'foundland, and at different times in our winter quarters the box-chronometers 

 were rated by Mr. Sonntag by means of a transit instrument. The mean rate of the pocket-chronometer 

 as found by comparison with each box-chronometer was adopted. As an approximate longitude of the 

 prime meridian of Rensselaer Harbor, T0° 40' W. of Greenwich has at present been adopted. A slight 

 change is anticipated from some observed occultations of planets by the moon and a solar eclijise : 

 these observations have not yet been worked up. Any change made hereafter in this longitude will, as 

 has already been remarked, equally affect all the other longitudes. 



" For the determination of azimuths by means of a sextant, the angle between the sun's centre and the 

 object was measured and the time noted. For this purpose the smaller telescope was used, and some- 

 times a pocket-sextant. Whenever the object, the azimuth of which was to be found, was farther 

 removed from 120° than from the sun, the angular distance of an intermediate object, about 90° from 

 the sun, was introduced. At the same time the altitude of the sun was observed, to allow for the 

 reduction of the arc of the horizon ; this reduction was always small, since the sun was seldom higher 

 than 30°, and in no case higher than 36°. 



" When the azimuth of an object was thus determined, a number of other conspicuous objects were 

 connected with it by horizontal angles. Two determinations of the azimuth of an object, obtained from 

 two astronomically-determined points, seldom differed more than seven minutes. 



