In connection with telescopes certain devices have from time to time been introduced, but these merely aim at the _convenience_ of the observer and do not supplant the broad principles upon which are based the various types of instrument above described. Such, for instance, are the "Siderostat," and another form of it called the "Coelostat," in which a plane mirror is made to revolve in a certain manner, so as to reflect those portions of the sky which are to be observed, into the tube of a telescope kept fixed. Such too are the "Equatorial Coudé" of the late M. Loewy, Director of the Paris Observatory, and the "Sheepshanks Telescope" of the Observatory of Cambridge, in which a telescope is separated into two portions, the eye-piece portion being fixed upon a downward slant, and the object-glass portion jointed to it at an angle and pointed up at the sky. In these two instruments (which, by the way, differ materially) an arrangement of slanting mirrors in the tubes directs the journey of the rays of light from the object-glass to the eye-piece. The observer can thus sit at the eye-end of his telescope in the warmth and comfort of his room, and observe the stars in the same unconstrained manner as if he were merely looking down into a microscope. Needless to say, devices such as these are subject to the drawback that the mirrors employed sap a certain proportion of the rays of light. It will be remembered that we made allusion to loss of light in this way, when pointing out the advantage in light grasp of the Herschelian form of telescope, where only _one_ reflection takes place, over the Newtonian in which there are _two_. It is an interesting question as to whether telescopes can be made much larger. The American astronomer, Professor G.E. Hale, concludes that the limit of refractors is about five feet in diameter, but he thinks that reflectors as large as nine feet in diameter might now be made. As regards refractors there are several strong reasons against augmenting their proportions. First of all comes the great cost. Secondly, since the lenses are held in position merely round their rims, they will bend by their weight in the centres if they are made much larger. On the other hand, attempts to obviate this, by making the lenses thicker, would cause a decrease in the amount of light let through. But perhaps the greatest stumbling-block to the construction of larger telescopes is the fact that the unsteadiness of the air will be increasingly magnified. And further, the larger the tubes become, the more difficult will it be to keep the air within them at one constant temperature throughout their lengths.