6 THE ROYAL SOCIETY OF CANADA 



During the season it was our plan to observe whenever weather 

 conditions would allow on the 16 principal compass points with both 

 port and starboard swings; and if it was only possible to complete a 

 swing with one helm, at the next opportunity a swing was made with 

 the other helm and the means of the positions and values of the con- 

 stants observed were used for the reduction of the observations. 



The method emplo5''ed in determination of the constants B, C, D 

 and E from the Deviations on port and starboard helm from the obser- 

 vation of August 3 and 4 is shewn below and following it is given a 

 summary of the values of the constants with the values of H and I 

 for the different locations and the calculation of the constants x, y, x' 

 and y' therefrom. 



These results may be exhibited graphically and it is in this manner 

 that the}^ are most useful. If the values of the constants for the Ritchie 

 compass for Latitude 57° 31' N, Longitude 59° 57' W and also for 

 Latitude 59° 44' N, Longitude 85° 57'"W just determined be now used 

 to determine the deviation on the different points and if these deviations 

 so determined are plotted on squared paper in such a way that 1 division 

 of abscissse represents 1 degree of deviation, — westerly and + easterly, 

 and 1 division of ordinate represents a change of 22^° in the direction 

 of the ship's head, and smooth curves be drawn through the different 

 points for the different positions, then the accompanying curves will 



be the result, where the one marked X X is the one for the first 



position and the one marked is the second. The effect produced 



on the deviations by change of magnetic latitude is quite pronounced 

 and in an iron ship the effect would be much greater, and this therefore 

 explains the need of navigators being furnished with charts of Inclina- 

 tion and Horizontal Force as well as Declination. 



The deviation curve having been found for the magnetic latitude 

 the ship is in, the deviation for any heading of the ship may be read off 

 the curve and applied to the compass course for steering purposes, 

 or as in our work it may be applied to the compass bearing of the sun 

 to give the true Magnetic bearing and then from the true bearing of the 

 sun the Magnetic Declination as shown in the example already given. 



The results of the ocean observations reduced to the mean of the 

 day are given below in table II. 



Land observations were made at Quebec, Rimouski, Battle Harbour, 

 Mar])le Island, Churchill, York Factory, Port Laperriere, Nottingham 

 Island, Sugluk Bay, Ashe Inlet, Stupart Bay and Port Burwell and the 

 results are tabulated and given below in table III. 



At Marble Island the time of a single oscillation of a magnet whose 

 moment was 280 dvnes was 8.7 seconds and its deflection of another 



