ON THE KEW MAGNETOGRAPHS. 329 



The same observations which determine the magnifying power of the lens, 

 afford also a means of estimating the amount of its spherical aberration. For 

 this purpose we have merely to examine whether the value of the apparent 

 scale in terms of the real scale remains constant at different portions of the 

 field. Observations of this kind having been made for all the instruments, 

 there is no reason to believe that, within the adopted range of motion, any 

 irregularity of scale exists. The scale divided on ground glass, which is 

 fixed permanently in each of the sliding plate-frames, supplies a means of 

 observing the positions of the magnets when the photographic registration 

 is not in action, and has been found of essential service in the preliminary 

 adjustments of the instruments. The scales are of the same value as that 

 employed in tabulating the numerical results : the zero division of the scale 

 can also be adjusted to correspond with the edge of the spot of light which 

 gives the zero-line on the plate ; observations taken in this way are there- 

 fore at once comparable with the photographic measurements. 



The length of time during which the light is allowed to act upon the plate 

 when the instrument is in action, is found by measuring the width of the dia- 

 phragm at the image end of the camera; or more accurately by measuring 

 the length of the impression produced by the light from the moveable slit 

 when the plate is at rest. This was found to correspond to about &\ mi- 

 nutes. 



The series of registrations by this instrument has been on several occa- 

 sions interrupted by derangements in the adjustments : these have been 

 shown in alterations of the relative positions of the fixed and moveable shields. 

 The moveable shield ought always to overlap, by a small quantity, the fixed 

 one, so as to stop all light except such as passes through the slits. It has 

 several times occurred that the shields have become so far separated as to 

 allow the light to spread across the plate and thus to prevent the appearance 

 of the curves. At other times the contrary has happened, the moveable 

 shield so much overlapping the fixed one as to obliterate the light from the 

 slit which produces the zero-line, the curve line being however still recorded. 

 On one occasion, about the end of July, it was found that the fixed shield 

 had become too high relatively to the image-diaphragm. For some days 

 previous to this being discovered, the registers were very faint, owing to the 

 great loss of light arising from this cause. These errors were always removed 

 by opening up the instrument and readjusting the parts deranged — an ope- 

 ration which always renders doubtful the connexion between the observa- 

 tions before and after the adjustments. The derangements now stated have 

 been ascribed to alterations in the condition of the wooden supports of the 

 instrument, arising from changes of temperature or humidity. They may 

 also be partly owing to the expansion and contraction of the very long sus- 

 pending string. 



Horizontal-Force Magnetograph. — The value in arc of the ordinates for 

 this instrument was found in the manner already described for the declina- 

 tion. The radius of the moveable shield ^9"08 inches, and the magnifying 

 power of the lens = 3*46 ; whence the arc-value of jJ^th of an inch=2'*188. 

 The arc-value was also determined on September 20th, 1850, by a different 

 process, as follows. The magnet having been removed, an equal weight was 

 suspended from the stirrup. There being then no magnetic directive force, 

 and the torsion-force of the bifilar suspension being considerable, when the 

 arms of the torsion-circle are turned through any angle, the arm carrying the 

 shield should move through the same angle. The image of the slit being 

 observed upon a divided scale placed at the focus, its motions corresponding 

 to certain changes of the circle reading were observed. From the mean of 



