The diameter of a star of the seventh magnitude as it appears in the focus of this huge telescope is 1 / 2,500 inch. The spiders' webs stretched across the object glass are about 1 / 6,000 inch in diameter. "The problem thus is," says Mr. Maw, "to move this twenty-two ton mass (the telescope) with such steadiness in opposition to the motion of the earth, that a star disc 1 / 2,500 inch in diameter can be kept threaded, as it were, upon a spider's web 1 / 6,000 inch in diameter, carried at a radius of thirty-two feet from the centre of motion. I think that you will agree that this is a problem in mechanical engineering demanding no slight skill to solve; but it has been solved, and with the most satisfactory results." The motions are controlled electrically; and respecting them Professor Barnard, one of the chief observers with this telescope, some time ago wrote as follows: "It is astonishing to see with what perfect instantaneousness the clock takes up the tube. The electric slow motions are controlled from the eye end. So exact are they that a star can be brought from the edge of the field and stopped instantaneously behind the micrometer wire."
Dividing engines are used for ruling parallel lines on glass and metal, to aid in the measurements of microscopical objects or the wave-lengths of light. A diffraction grating, used for measuring the latter, has the lines so close together that they would be visible only under a powerful microscope. Glass being too brittle, a special alloy of so-called speculum metal is fashioned into a highly polished plate, and this is placed in the machine. A delicate screw arrangement gradually feeds the plate forwards under the diamond point, which is automatically drawn across the plate between every two movements. Professor H. A. Rowlands has constructed a parallel dividing engine which has ruled as many as 120,000 lines to the inch. To get a conception of these figures we must once again resort to comparison. Let us therefore take a furrow as a line, and imagine a ploughman going up and down a field 120,000 times. If each furrow be eight inches wide, the field would require a breadth of nearly fourteen miles to accommodate all the furrows! Again, supposing that a plate six inches square were being ruled, the lines placed end to end would extend for seventy miles!
Professor Rowlands' machine does the finest work of this kind. Another very perfect instrument has been built by Lord Blythswood, and as some particulars of it have been kindly supplied, they may fitly be appended.
If a first-class draughtsman were asked how many parallel straight lines he would rule within the space of one inch, it is doubtful whether he would undertake more than 150 to 200 lines. Lord Blythswood's machine can rule fourteen parallel lines on a space equivalent to the edge of the finest tissue paper. So delicate are the movements of the machine that it must be protected from variations of temperature, which would contract or expand its parts; so the room in which it stands is kept at an even heat by automatic apparatus, and to make things doubly sure the engine is further sheltered in a large case having double walls inter-packed with cotton wool.
In constructing the machine it was found impossible, with the most scientific tools, to cut a toothed wheel sufficiently accurate to drive the mechanism, but the errors discovered by microscopes were made good by the invention of a small electro-plating brush, which added the thinnest imaginable layer of metal to any tooth found deficient.
During the process of ruling a grating of only a few square inches area, the machine must be left severely alone in its closed case. The slightest jar would cause unparallelism of a few lines, and the ruin of the whole grating. So for several days the diamond point has its own way, moving backwards and forwards unceasingly over the hard metal, in which it chases tiny grooves. At the end the plate has the appearance of mother-of-pearl, which is, in fact, one of nature's diffraction gratings, breaking up white light into the colours of the spectrum.
You will be able to understand that these mechanical gratings are expensive articles. Sometimes the diamond point breaks half-way through the ruling, and a week's work is spoilt. Also the creation of a reliable machine is a very tedious business. Ten pounds per square inch of grating is a low price to pay.
The greatest difficulty met with in the manufacture of the dividing engine is that of obtaining a mathematically correct screw. Turning on a lathe produces a very rough spiral, judged scientifically. Some threads will be deeper than others, and differently spaced. The screw must, therefore, be ground with emery and oil introduced between it and a long nut which is made in four segments, and provided with collars for tightening it up against the screw. Perhaps a fortnight may be expended over the grinding. Then the screw must undergo rigid tests, a nut must be made for it, and it has to be mounted in proper bearings. The explanation of the method of eliminating errors being very technical, it is omitted; but an idea of the care required may be gleaned from Professor Rowlands' statement that an uncorrected error of 1 / 300,000 of an inch is quite sufficient to ruin a grating!
In the Houses of Parliament there is kept at an even temperature a bronze rod, thirty-eight inches long and an inch square in section. Near the ends are two wells, rather more than half an inch deep, and at the bottom of the wells are gold studs, each engraved with a delicate cross line on their polished surfaces. The distance between the lines is the imperial yard of thirty-six inches.