762 Description of a Persian Astrolabe. j^No. 118. 



also passing through the centres of the Planisphere discs, together with 

 the trelliced plate (Plate II) on the face, binds the whole firmly and com- 

 pactly. The observer now suspends the instrument, by holding the 

 ring mentioned above in his right iiand, the line passing through the 

 origin of the graduated arcs, and the centre thus being horizontal, 

 while that through their terminus, and at right angles to the former, 

 is necessarily vertical. The object to be observed is now made to 

 coincide with the plane of the quadrants, and the label is turned 

 towards it, until a ray of light pass from it to the eye of the observer, 

 through the perforations of two small plates projecting from the label 

 near its ends, and at right angles to its length, (Plate I. Fig. b,) The 

 arc of the quadrant between the horizontal line and the edge of the 

 label, will evidently be the altitude required. There are undoubtedly 

 several sources of error, which would render such an instrument unfit 

 for the purposes of modern science, but with those for whose use it 

 was originally intended, its imperfections would be unimportant. It 

 is evident, that in the observation described, the altitude obtained is 

 too great by the whole amount of refraction, an error which becomes 

 considerable, when the object is near the horizon. I am of opinion, to, 

 that the error in reading off" the altitude would not be sufficiently 

 allowed for, when repeated observations are impracticable, under 5.' 

 Another ground of inaccuracy would necessarily be the difficulty, almost 

 impossibility, with such an instrument of taking the centre in observa- 

 tions of the sun. These errors, springing from different sources, might 

 sometimes, it is true, correct each other to a certain extent, but this 

 vague probability must, of course, be insufficient to produce confidence 

 in the instrument. 



The parallel straight lines on the left of the upper limb are semi- 

 almacanthers, or semi-circles of celestial altitude, seen on their edges, 

 while the concentric arcs on the right seem to be intended to connect 

 those signs which have north declination with those which, in that 

 respect correspond with them towards the south ; such a table is 

 committed to memory by Asiatic astronomers, to which the neat 

 arrangement is well suited. 



Under the central line, and symmetrical with the centre, are two 

 rectangles, the one within the other, and whose length is twice their 

 breadth. The figures alluded to, are immediately recognizable by 



