452 Sir W. Herschel’s astronomical observations 
supplied in a figure representing such a globe, by the addition 
of lines that are of a length which is proportional to the 
diameter of it. 
It has been shown in my last paper, that all the stars which 
may be seen in the clearest nights, are probably contained 
within a globular space, of which the radius does not exceed 
the 12th order of distances ; I shall therefore suppose the 
circle c in the centre of fig. 1, [Plate XXI.] to represent a 
celestial globe, containing all the stars that are generally 
marked on its surface ; their arrangement within this globular 
space, however, must be supposed to be according to their 
order of distances, the stars of the first order being placed 
nearest the centre, and those of the 2d, 3d, and 4th, &c. 
gradually farther off ; but they must all be placed in their 
well ascertained directions, so that a line from the centre 
drawn through any one of them may come to the surface at 
the place where its situation is marked. 
According to this assumption it follows, that all those 
celestial objects which are farther than the 12th order of 
distances from the centre, must be represented as being at 
the outside of the globular space ; but as our celestial globes 
represent not only the situation of the stars of the heavens, 
but give us also many additional objects, such as clusters of 
stars, nebulas, and the milky way, it is evident that the point 
where the line of sight from the centre to any one of these 
distant objects, leaves the surface of the globular space, is 
ascertained ; and since any celestial object not inserted on 
our globes, of which the right ascension and polar distance 
are given, may be easily added, the position of the visual ray 
directed to such an object will thereby also be determined. 
