TERRESTRIAL MAGNETISM. 697 



fibre of flax, also by finding in what direction a given number of 

 vibrations was made in the shortest time, but no result could b3 

 obtained. Six sets of observations were taken in what seemed to 

 be the most probable direction, and at 45° and 90° to it, and the 

 mean of all the observations gave the dip 89° 59'. 



The latitude of the position was 70° 5' 17" N. 



The longitude „ „ 96° 45' 48'' W. 



6. Magnetic Observations in the Arctic Seas, by Dr. E. K. 

 Kane, during the Second Grinnell Expedition in 1853- 

 65, at Van Rensselaer Harbour. (" Smithsonian Con- 

 tributions," vol. X.) 



Observations for changes of magnetic declination were made 

 with a unifilar magnetometer. The magnet was suspended by a 

 silk string 9^ inches in length, and on several trials the effect of 

 torsion was found to be exceedingly small. The Expedition not 

 being well provided with instruments, the same instrument was 

 employed for observing absolute declinations. Mr. Sonntag ob- 

 tained a few values in winter quarters by detaching the box con- 

 taining the magnet from the circle which bears the telescope. The 

 telescope was then moved in azimuth until a well-defined object 

 within the range of its vertical motion could be observed. The 

 same instrument was also employed to determine the Earth's 

 horizontal magnetic force. 



The instrument was perched on a pedestal of frozen gravel, 

 the contents of two barrels, and this mounting was as stable as 

 the rock underneath. 



At Van Rensselaer Harbour the Observatory was placed upon 

 the northernmost of the rocky group of islets that formed the 

 harbour. It is 76 English feet from the highest and northern- 

 most salient point of this island, in a direction S. 14° E., or in one 

 with said point and the S.E. projection of the southernmost islet 

 of the group. A natural face of gneiss rock formed the western 

 wall of the Observatory. A crevice in this rock had been filled 

 with melted lead, in the centre of which is a copper bolt. Eight 

 feet from this bolt, and in the direction indicated by the crevice, 

 stood the magnetometer. This was called the Fern Rock Obser- 

 vatory. The highest point of the island was about 30 feet above 

 the mean tide level of the harbour. 



On the 9th of June 1854 Mr. Sonntag examined the instru- 

 ment in reference to local disturbance, and found no sensible 

 deviation arising from such a source. " The local deviation seems 

 " to have corrected itself ; the iron in our comfortless little cell 

 " seems to have been so distributed that our results were not 

 " affected by it." Hourly observations for diurnal vai-iation were 

 made during January, February, and March 1854, generally at 

 Intervals of three or four days, as well as on term days. The 

 term days were January 18, 19, February 24, 25, INFarch 22, 23, 

 April 19, 20, May 26, 27, June 21, 22, 1854. Readings (the 

 mean of two extremes during a vibration when the magnet is in 

 motion) were taken every sixth minute on term days, beginning 



