356 
ME. EVAJs’S OX THE DEVIATIOX OF THE COMPASS 
the latter on the compass in heeling may be gathered, and the corresponding effect 
shown, under the various directions of the ship’s head in building : thus — 
With head built North, on heeling, the north end of compass needle will be attracted 
to the weather or nearest side from its south polarity. 
,, „ the same. 
„ „ the same. 
,, the north end of needle uill have but little error from 
the balanced conditions of north and south polarity 
of topsides. 
„ the north end of needle will be repelled to the lee 
side by the north polarity of nearest or weather top- 
side. 
S.W., „ the north end of needle will have but little error, as 
at S.E. 
West, „ the north end of needle will be atti'acted to the 
weather or nearest side. 
N.W., „ „ the same*. 
N.E., 
East, 
S.E., 
South, 
These laws only hold good as long as the topsides iir the immediate vicirrity of the 
compass retain their dominant polarity due to their original direction of birild in Great 
Britairr : if in south magnetic latitudes a change of polarity takes place, the conditions 
of heelirrg correspond to such change. 
The maximum disturbance on heeling in all these vessels is when their heads ar'e (by 
disturbed compass) magnetic North or South, aird this distirrbarrce varrishes when the 
head is East or West. This law of disturbance may be thirs explained : when the vessel’s 
head is north or south on an everr keel (by disturbed compass), the rreedle lies par-allel 
to the topsides by their combined actiorr, which rreirtralizes each other ; orr heelirrg, the 
nearest topside exercises its then domirrairt polarity at right arrgies to the directiorr of 
the rreedle, and hence the maximum error. With the ship’s head east or west, whether 
on an even keel or heeling, either pole of the compass-needle poirrts directly to the 
topsides, and is consequently unaffected except hr a vertical plarre. 
As the amount of disturbance orr heelirrg varies rrrrder the varioirs couditiorrs of direc- 
tion of build, height of compass, arrd breadth of ship or distairce of topsides, added to 
the prevailing permanent or iirductive rnagrretic corrditiorr of the latter arrd the deck 
beams, each ship must have arr individrral character, to be determirred only by experiment 
or observation at sea. There are, however, stroirg grounds for irrferriirg that by a 
* These principles have in their main features been confirmed by the experiments and investigations of 
the Secretary of the Liverpool Compass Committee, — Mr. Eundell, — to whom I am indebted for much 
valuable information on many questions of theoretical and practical value ; but their simple illustration by 
a consideration of the appended diagrams, which, so far as I am aware, are novel in their application, will 
probably be found instructive and useful to the practical seaman. 
