24 , REPORT 1860. 



perhaps, is to have the bar hollow, and to employ reflexion of a cross wire from a 

 surface of mercury. The bar is then moved in the magnetic meridian from the ver- 

 tical position till the deflection of the magnet is zero ; if the permanent magnetism 

 acts with the induced magnetism, the movement of the bar will be greater than the 

 inclination by a given angle ; in turning the bar in the opposite direction (so as to 

 invert it), the angle from the vertical will be less by the same amount. 



Since in the position at right angles to the magnetic force the induced magnetism 

 is zero, the objection applying to Dr. Lloyd's method does not exist ; there is, how- 

 ever, still a source of error remaining that applies to both : as the magnetic inclina- 

 tion increases, the position at right angles to the force can only be attained by moving 

 the bar nearer and nearer to the horizontal, and as it approaches the horizontal, a 

 certain amount of magnetism is induced in the bar by the small suspended magnet. 

 Different methods have been imagined by the author to destroy or balance this 

 action ; but the best method he thinks will be to make observations with the bar at 

 two different distances. The magnetism induced by the small magnet in the bar 

 may be represented by a weak magnet, whose force will vary inversely as the cube of 

 the distance : as the action of this weak magnet will also vary inversely as the cube 

 of the distance, the effect may be determined and eliminated by observations at two 

 or more distances. 



Any error of the observation for the vertical position of the bar due to the non- 

 coincidence of the axis of magnetism and of figure may be eliminated by turning the 

 bar on its vertical axis of figure through 180°. 



The author remarked that the error due to the inducing action of the small 

 suspended magnet might be rendered as small as we please, by employing a modi- 

 fication of the method used by him in India. If the total deflection due to the bar 

 vertical (direct and inverted) be determined, and we then observe the change of de- 

 flection due to a given angular movement of the bar from the vertical, we may com- 

 pute the movement necessary to render the deflection zero : the angular movement 

 may be taken of such magnitude as to render the effect of the inducing action negli- 

 gible. This modification requires, however, the determination of the angles of 

 deflection, and therefore is far from the simplicity of the first method. The author 

 pointed out, that, since when the bar is at right angles to the direction of the total 

 force any small movement of the bar will produce induced magnetism in proportion 

 to the sine of the small angle of movement, this position is that best fitted to give 

 the true position of the magnetic meridian with the least error of inclination. 



The author concluded by stating that he had learned since his return to Europe 

 that Dr. Lamont had also proposed a method differing from that of Dr. Lloyd. Dr. 

 Lamont employed an astatic needle, and turned the bars into different azimuths by 

 movement on a vertical axis, so as to produce different amounts of induced magnet- 

 ism without changing the position of the bars relatively to the vertical. This 

 method, Dr. Lamont informed the author, had failed on account of the bars receiving 

 different amounts of permanent magnetism in changing from azimuth to azimuth. 

 This difficulty does not exist in Mr. Broun's method, as the bar is always kept in the 

 meridian, and is always brought to the position where the inducing action is zero. 



On Magnetic Rocks in South India. By John Allan Broun, F.R.S. 



The Moocoonoomalley is a granite hill rising about 800 feet above the sea, 5 miles 

 south-east of Trevandrum, and about 35 miles north-west of Cape Comorin. General 

 Cullen, the late British Minister at Travancore, had observed several anomalies in the 

 magnetic dip in ascending this hill. The dip near Trevandrum and about the base 

 of the hill was from 2° 30' to 2° 40' S. ; on the top he found the dip to be from 

 5° 52' to 11° 23' in different years, in which he probably slightly varied the position 

 of observation. 



In December 1855 I examined the rock masses constituting the hill. The plain 

 around the base is formed of a stratified rock known to Indian geologists by the 

 name of laterite. The first rocks in the ascent are dark syenites, containing a con- 

 siderable proportion of hornblende (in some cases the appearance is more like a 

 greenstone) ; towards the middle of the ascent light-grey syenites become common, 



