345 
Mr Barlow’s Discoveries in Magnetism, 
fitting to calculation the local effect of a ship’s guns on the 
compass. He began his observations, on the effect of balls 
of different dimensions, and in the course of these, he. 
was led to the following curious experimental fact, viz. that 
there is round every globe and mass of iron a great circle in- 
clining from north to south, and forming with the horizon an 
angle of about 19 or SO degrees, in the plane of which the 
iron has no effect in changing the direction of the needle, that is 
to say, while the centre of the needle is found in the plane of 
this circle, the compass will point north and south, the same as 
if no iron were present. 
The dip of the needle being about 70° 30', he apprehended 
that the inclination of this circle was equal to the complement 
of the dip ; and subsequent experiments on an excellent hori- 
zontal compass and dipping needle, furnished by Mr Berge, 
have confirmed the accuracy of these surmises. 
This fact being established, his next object was to ascertain 
what law was observed in the attraction of iron when the com- 
pass was removed out of the above circle of no attraction ; and 
with this view, he contrived, by means of an apparatus con- 
structed for the purpose, to carry the compass round the ball 
(which was 13 inches in diameter, and solid, weighing S88 lb.) 
in various circles ; and by these means succeeded in deducing a 
law of action, which was singularly uniform, the computed and 
observed results scarcely ever deviating from each other, by a 
quantity greater than the daily variation, viz. from 10 to 20, or 
at most 30 minutes. 
The nature of the above laws will be best conceived by the 
following artificial consideration. Call the circle of no attrac- 
tion above described, the magnetic equator of a sphere circum-- 
scribing the ball^ and its two poles, the poles of that sphere. Con- 
ceive now circles of latitude and longitude to be drawn, the first 
meridian of the latter passing through the east and west points of 
the horizon, and the magnetic equator. Then the law in question 
is this. That the tangent of the deviation of the needle from 
the north or south, is proportional to the rectangle of the sine of 
the double latitude and cosine of the longitude ; so that, know- 
ing the deviation in one instance only, it may be computed for 
any other whatever. 
