this point the difficulties begin.  For speculum metal when cold is excessively brittle, and were the casting permitted to cool like an ordinary copper or iron casting, the mirror would inevitably fly into pieces.  Lord Rosse, therefore, found it necessary to anneal the casting with extreme care by allowing it to cool very slowly.  This was accomplished by drawing the disc of metal as soon as it had entered into the solid state, though still glowing red, into an annealing oven.  There the temperature was allowed to subside so gradually, that six weeks elapsed before the mirror had reached the temperature of the external air.  The necessity for extreme precaution in the operation of annealing will be manifest if we reflect on one of the accidents which happened.  On a certain occasion, after the cooling of a great casting had been completed, it was found, on withdrawing the speculum, that it was cracked into two pieces.  This mishap was eventually traced to the fact that one of the walls of the oven had only a single brick in its thickness, and that therefore the heat had escaped more easily through that side than through the other sides which were built of double thickness.  The speculum had, consequently, not cooled uniformly, and hence the fracture had resulted.  Undeterred, however, by this failure, as well as by not a few other difficulties, into a description of which we cannot now enter, Lord Rosse steadily adhered to his self-imposed task, and at last succeeded in casting two perfect discs on which to commence the tedious processes of grinding and polishing.  The magnitude of the operations involved may perhaps be appreciated if I mention that the value of the mere copper and tin entering into the composition of each of the mirrors was about 500 pounds.

In no part of his undertaking was Lord Rosse’s mechanical ingenuity more taxed than in the devising of the mechanism for carrying out the delicate operations of grinding and polishing the mirrors, whose casting we have just mentioned.  In the ordinary operations of the telescope-maker, such processes had hitherto been generally effected by hand, but, of course, such methods became impossible when dealing with mirrors which were as large as a good-sized dinner table, and whose weight was measured by tons.  The rough grinding was effected by means of a tool of cast iron about the same size as the mirror, which was moved by suitable machinery both backwards and forwards, and round and round, plenty of sand and water being supplied between the mirror and the tool to produce the necessary attrition.  As the process proceeded and as the surface became smooth, emery was used instead of sand; and when this stage was complete, the grinding tool was removed and the polishing tool was substituted.  The essential part of this was a surface of pitch, which, having been temporarily softened by heat, was then placed on the mirror, and accepted from the mirror the proper form.  Rouge was then introduced as