The next point worthy of attention is the question of permanence of optical qualities. Here the refractor undoubtedly has the advantage. It is true that the flint glass of some objectives gets attacked by a sort of tarnish, still, that is not the case generally, while, on the other hand, metallic mirrors often become considerably tarnished after a few years of use, and although repolishing is not a matter of any great difficulty in the hands of the maker, still it is a serious drawback to be obliged to return mirrors every few years to be repolished. There are, however, some exceptions to this, for there are many small mirrors in existence whose polish is good after many years of continuous use, just as on the other hand there are many object-glasses whose polish has suffered in a few years, but these are exceptions to the rule. The same remarks apply to the silvered glass reflectors, for although the silvering of small mirrors is not a difficult process, the matter becomes exceedingly difficult with large surfaces, and indeed at present large discs of glass, say of four or six feet diameter, cannot be produced. If, however, a process should be discovered of manufacturing these discs satisfactorily and of silvering them, there are objections to them on the grounds of the bad conductivity of glass, whereby changes of temperature alter the curvature to a fatal extent, and there is also a great tendency for dew to be deposited on the surface. The next point to be considered is the general suitability for observatory work, and this depends upon the quality of the work required, whether for measuring positions, as in the case of the transit instrument, where permanency of mounting is of great importance, or for physical astronomy, when a steady image for a time is only required. For the first purpose the refractor has decidedly the advantage, as the object-glass can be fixed very nearly immovably in its cell, whereas its rival must of necessity, at least with present appliances, have a small, yet in comparison considerable, motion. Again, the refractor has the advantage over the other in not being of so large aperture when of equal power, so that the disturbing effects of air currents is considerably less, but the method of making the tubes of open lattice-work materially reduces this objection. We have mentioned the difficulty of mounting mirrors, especially of large size, but this has now been got over very perfectly. This difficulty does not occur in the mounting of object-glasses of sizes at present in use, but when we come to deal with lenses of some 30 inches diameter, the present simple method will in all probability be found insufficient. On the other hand the cost of mirrors is of course much less than that of object-glasses, a matter of considerable importance. The late M. Merz, on being asked as to price of a 30-inch object-glass, estimated that, if it were possible to make it, its cost would be between £8,500 and £9,000.