14 
MR. BARLOW ON THE PERFORMANCE 
case of open lenses), as also indirectly from principles of his own. I have not 
as yet had an opportunity of examining these cases, but, from the well-known 
ingenuity of their author, I cannot doubt of finding in this memoir many 
useful suggestions. 
The great change, however, which I propose to make in the construction of 
this giant telescope, is to have two front lenses, which will be attended with 
advantages not involved in the above considerations. At present, in conse- 
quence of the diameter of the fluid lens being only half that of the front lens, 
it is difficult to get a sufficient quantity of spherical aberration in the former, 
to correct that of the latter ; — for although we give to the plate lens the cur- 
vature requisite for reducing its aberration to a minimum, yet the fluid lens is 
obliged to be made considerably concavo-convex (a form not to be used when 
it can be avoided), in order to produce a sufficient aberration in the fluid to 
correct it. Moreover I have hitherto employed parallel meniscus cheeks to 
contain the fluid, which present a practical difficulty, if not a positive impedi- 
ment, to good centering. This will be seen immediately when we consider 
that when a lens is double-concave or convex, and also when it is concavo- 
convex, if the radii of curvature are very unequal, the centering may always 
be effected : for the line joining the centres of the two spheres, or this line pro- 
duced in the latter case, must pass through the lens, and indicate its true 
centre : but when the lens has parallel surfaces, or the radii equal, if the two 
spherical centres be not coincident from the tool itself (a very improbable case) 
the line which joins them can never cut the lens, and consequently it can have 
no true centre. All these evils will, however, be avoided in the proposed 
application of two front lenses, which, by being placed each in what opticians 
call their best position, will at once reduce the spherical aberration of the 
front lens to about one third of its present amount, and thereby enable us to 
correct it by the fluid lens without adopting the distorted form rendered 
necessary under present circumstances. 
Another important consideration is also involved in this form, relative to the 
facility it affords of obtaining the plate-glass. If the front lens were single, 
the thickness would be such as would require the glass to be made specifically 
for the purpose, and of course all the delay and expense of previous experi- 
ments would be incurred ; whereas, by dividing the whole amount of curvature 
