268 Scientific Proceedings, Royal Dublin Society. 
are altogether in air; and if the mirror were supported only by its 
edge, it would sag down when pointed towards the sky, the central 
ray would then have farther to travel 
to reach the mirror than the marginal m 
rays, and again farther to get from the 
mirror to 0. The wave front of the 
central ray will accordingly reach o Be) ig. 
too late to unite with the waves 
travelling along a ore. These will already have passed o. 
Hence a speculum must be kept from sagging: and to this end 
must be supported continuously over its whole back, and with such 
extreme delicacy that all parts of the back of the mirror shall be 
equally pressed. 
If the mirror could be kept horizontal, it might be floated on 
mercury; but as during the use of the telescope it has to be sloped 
into different positions, a layer of the delicate springs of which 
flannel consists has been used for the support of smaller specula 
up to a 15” or 18” aperture, and the mechanical contrivance called 
a bed of levers for the larger ones, the edge of the mirror being 
supported in some way which allows its back to lean freely against 
the bed of levers, or layers of flannel, as the case may be. 
Whether it be the levers or the layers of flannel that are used, 
they inevitably yield when the pressure upon them varies, and it 
becomes difficult to prevent the line of collimation—that is, the 
optical axis of the mirror—from shifting a little relatively to the 
tube of the telescope, when the telescope is carried from one posi- 
tion to another—so difficult that it is only in the hands of very 
skilful manipulators, like Dr. Isaac Roberts, that celestial photo- 
graphs can be satisfactorily taken with reflecting telescopes. Never- 
theless, when they are so taken, they surpass those which ean be 
produced by refractors. 
2.—THE Proposep Mountine. 
The following seems to be a way in which this difficulty can 
be entirely gotten rid of, and specula made available both for 
celestial photography and spectroscopy, and in astronomical in- 
struments of precision :— 
Let the mirror If be made the front of an airtight cell, such as 
that represented in the diagram, which consists of three chambers 
