Page:The American Cyclopædia (1879) Volume XV.djvu/267

SPECULUM 255 sists of copper and tin, to which Newton added a little arsenic, and sometimes silver or zinc ; but Lord Rosse found that the two metals first named are better without the addition of any other. The volatile metals are objectionable. Silver makes the alloy too soft ; nickel, though it whitens the yellow alloy of copper, makes the speculum alloy yellowish. Rosse was particu- lar that the copper and tin should be com- bined in their atomic proportions (4 atoms of copper =126-8, to 1 of tin =59), and the pur- est metals should be selected ; for the smaller specula it is even recommended that the cop- per be obtained by the electrotype process; this is hardly practicable for the large ones. The alloy is remarkable for its extreme brit- tleness and hardness. Large masses of it sometimes break from a slight blow or sud- den change of temperature ; and it is so hard that it cannot be wrought with tools of steel. It takes a most brilliant polish, which it has been known to retain with little tarnish, though exposed to the air for more than 16 years. A large speculum, however, ought al- ways to bo covered when not in use, and the air about it should be kept dry by means of an open box of quicklime. Great difficulties have been encountered in preventing the large specula from changing their form by their own weight, and those of 6 ft. diameter are made so thick, to give them the necessary stiffness (though supported when finished by the most ingenious appliances), that they are among the heaviest of bronze castings ; and the prepara- tion of the rough mass is among the most diffi- cult of foundery operations. Rosse's six-foot speculum weighs 4 tons ; one of 3 ft. 3 in. thick, weighs 13 cwt. ; and one of 2 ft., 3 cwt. The alloy is prepared by melting the metals separately, and pouring the tin into the copper, stirring rapidly, and then, before the tin oxi- dizes, casting the alloy into ingots. It is test- ed when cold to ascertain its brilliancy, and more tin is added if necessary. The best mode of preparing the moulds has been arrived at from long experimental trials in the casting of the smaller specula. It was found that sand moulding would not answer for the surface of the disks, as the texture of the alloy near the outside was rendered somewhat spongy and crystalline ; and though this was so slight as to be detected only by the microscope, it still seriously impaired the polished surface. The face of the disk at least must then be " chilled," as cast iron is chilled, by pouring it into metal- lic moulds to increase the density of its sur- face. But the ordinary temperature of the atmosphere was found to be too low for the moulds to receive this alloy, and they were consequently heated to about 212 to prevent too sudden cooling and consequent irregular contraction. For moderate-sized specula cast- iron moulds were used, necessarily open, or the casting would inevitably fly in pieces. They were made a little deeper than the spec- ulum, with the bottom of the same convexity 757 VOL. xv. 17 with this, and so supported that they could be instantly filled from the lowest point, and turned into a horizontal position when charged with the proper weight of the metal. The air and any foreign substance present are thus carried up to the surface, and separated from the alloy. But this was not sufficiently per- fect for the largest castings, and Lord Rosse adopted for these the following method. An iron frame of sufficient diameter was filled with pieces of hoop iron set on edge and tight- ly wedged together, and the upper surface was turned off to the curvature of the face of the speculum. This was to serve for the bottom of the mould, being tight enough to hold the melted metal, while it allowed the escape of the gases through the interstices uniformly over the whole face. Upon this bottom was laid the wooden pattern, made twice as deep as the intended speculum, and with an allowance of -5*5- in the diameter for shrinkage. The sides of the mould were then formed by ramming sand around the pattern. By this arrange- ment the first cooling is on the under face, next on the sides, and the final congealing is on the top or back, where the contraction and resulting irregularities will be concentrated in the least important part. The metal while red-hot is removed to a fuFnace specially pre- pared for it, the bottom having the curvature of the disk (unless in case of using an iron mould, when this too is taken along). By fires already kept up several days, the inner walls of the furnace should be at a full red heat. The vacant spaces around the casting are then filled with ignited fuel, and every aperture is care- fully luted. A large speculum should thus be left to cool for a month to six weeks; and the result may still be unsatisfactory if the walls of the furnace are less than 2 ft. thick. The production of the true parabolic figure, com- bined with a brilliant polish, is attained by grinding succeeded by polishing. Machines applicable to this object have been invented by Sir W. and Sir J. Herschel, Lord Rosse, Mr. Lassell, an amateur optician and astronomer, Mr. De la Rue, Mr. Grubb of Dublin, and others, which are too complicated to be par- ticularly described in this place. The object sought for is to restrict the operation of the rubbing tools to the production of the particu- lar curvature required, and insure a uniform action upon every part of the surface of the disk. The speculum, placed upon a slowly re- volving platform, presents its face to the action of the rubber above it, which by Lord Rosse's arrangement was caused to vibrate regularly in one and the other direction, while at the same time it revolved at a different rate from that of the speculum. Mr. Lassell caused the rubber to revolve in small circles, while the speculum, turning on its axis, which was not in line with that of the rubber, presented suc- cessively all portions of its surface to this circling action of the rubber, thus imitating the movements of the hands by which the