282 SUPPLEMENT TO THE 



tube in which they are fixed. This defect is easily known, 

 by observing the position of the luminous point in the 

 centre of the discs they generate, which, if perfectly cen- 

 trical, indicates perfect adjustment, while the degree of its 

 eccentricity denotes the magnitude of the error in this 

 particular : thus, (13) shews the phenomenon presented 

 by a triple object-glass thrown violently out of adjust- 

 ment ; and (b, 14), another disc, of a similar description, 

 but more out of focus. In observing for adjustment, it is 

 especially necessary that the object employed should be 

 placed exactly in the centre of the field of view, for it is 

 impossible that an object-glass can be rigorously adjusted 

 for more than one pencil. This may be either a direct or 

 an oblique one, as we please ; but it is impracticable to 

 make a perfect adjustment for one pencil consist with a 

 perfect adjustment of another; for whatever part of the 

 field of view the adjustment has been effected, there the 

 maximum of distinctness will be found. The state of ad- 

 justment may moreover be determined upon by using the 

 rings 5' on a dial-plate ; if it is perfect, it will be well re- 

 presented by 8 X// , where the penumbra goes off equally on 

 all sides of the ring alike, leaving a dark centrical spot, 

 and a true circular figure at the margins. The adjustment 

 of metals with each other is denoted by precisely the same 

 phenomena as that of object-glasses, always allowing for 

 the effect of the shadow of the small metal and its arm. 



N.B. The indications given by large globules of quick- 

 silver frequently differ widely from those given by small 

 ones, in all respects. We may term a globule large when 

 it is visible to the naked eye, as at (5) ; small, when it is 

 not. Thus, a large globule, giving an image of a window 

 in focus, will, when put out, present only a disc, like (9), 



