PASSAGE FKOM GREENLAND TO ErxITAlN. 377 



equal error in the same direction when sailing on 

 an east course ; so that when a ship's head, by the 

 compass placed near the middle of the quarter-deck, 

 appears to be west-, it is actually W. ^ S. or W. ^ S, 

 compared with the true magnetic meridian ; and 

 when it appears to be east, it is E. ^ S. or E. ^ S. 

 On more oblique courses, such as N. E., S. E., 

 S. W. and N. ^.V. the equation or anomaly of va- 

 riation^ though less considerable, is still very per- 

 ceptible. A ship steering with a fair wind from 

 Greeenland, S. W. :|; W. by the compass, scarcely 

 makes good a S. S. W. course, though the variation 

 should be but two points. This quarter point of 

 attraction or anomaly of variation, all the way 

 from latitude 78° to 61®, a distance of 1020 miles, 

 occasions an error of 152' of longitude, or about 

 2 3 degrees, which the reckoning will be to tlic 

 westward of the ship *. This, of itself, is nearly 

 sufficient to account for the frequent mistakes Vvhich 

 are made, but amply so if we view it in connection 

 with the liability to err, occasioned by the insuffi- 

 ciency of the usual allowances made for lee-way, 

 when westerly winds prevail, together with the other 

 source of error already mentioned, in the departure 



* This subject, the anomaly of variation, having been con- 

 sidered more at large in the Fiist Volume of this Work, it is 

 unnecessary to state in this place the facts and observations 

 from which the above conclusions arc derived. 



