63 



REPORT 1872. 



These numbers, examined by difi'erences, indicate a formula whose principal 

 term involves the square of the temperature. If they be projected as ordinates, a 

 curve of parabolic form results, as given above, the vertex being- placed somewhere 

 below 40°, where, probably, a cune of contrary flexure would begin. Assuming 



the curve to be a parabola, ^ mnn ~^ *^® correction for temperature, and 



taking n-- 



1000 

 :10° and /3 = '0837, the values of e are given in the third column. 



A still closer approximation is found by employing a more complete expression 



CI 7^ T" I C T^ 



for the function of temperature, such as iTjnfF — ' ^"^ which, between 40° and 



80°, « = 0, 6= •OS, c = ^000415. The resulting corrections appear as inider, from 

 90° to 30°, the utmost limits of probable observation in the British Isles. 



A Table constructed in accordance with this formula has been found, by much 

 experience, to give very satisfactory results in measures of terrestrial elevation. 

 The author has also investigated instruments compensated for temperature, and 

 finds in some cases a curve-correction necessary for inequality of scale. 



Description of the new Marriotti Barometer, By Macneil Telford. 



The Spiral Top. By Prof. Ch. V. Zengee. 



■" On a stand of brass, fastened to a board, a screw moves in a vertical direction on 

 the upper end of the stand. The axis of the top, with its conical end, is put in a 

 conical hole worked in the screw, and the other end lies in a similar conical hole 

 on a support fastened to the board vertically and in the same line as the screw. 



On the axis of the top is fastened a movable arm, with a screw and tall of 

 brass that can be fastened at different distances from the axis of rotation. 



A double ring of brass to .'pan a sheet of paper in it may be fastened on the 

 screw, the plane of the paper being vertical to the axis of the top. Q'he apparatus 

 is then prepared to show the nutation of the axis and precession of the nodes 

 by spinning it. The axis of the top describes a circle on the paper, on which an 

 ellipse evolves, whose lengtli of axis depends on the position of the arm on the 

 axis and the distance of the ball screwed to it. Taking the brass arm with the 



