DIFFRACTION GRATING REPLICAS (Fr., Replicas des reseaux de diffraction; Ger., Digraktions-Gitter-Abklatsche) Original gratings, whether on glass or metal, are extremely expensive, and numerous attempts have been made by Rayleigh, Abney, and others to reproduce these by photographic means, but the most successful way is that of taking casts in celluloid. Thorpe, Ives, and Wallace produce these, and the latter has given full working instructions, of which the following is an abstract :— Pyroxyline . . 181 grs. 3.9 g.
Pure amyl acetate . r oz. TOO ccs.
Add the pyroxyline in small quantities to the amyl acetate, shake well till dissolved, and allow to stand for twenty-four hours. At the end of that time the resultant collodion should be poured from a height of at least 3 ft. or 4 ft. in a very thin stream into a large tray filled with water, the latter being constantly stirred with a glass rod. In about twenty-four hours the whole of the pyroxyline is precipitated in the form of white or light grey flocculent masses, which should be collected and dried. The pur pose of this precipitation is probably to purify the pyroxyline, as any collodion poured into water gives up what Eder has called " pyroxyline gum." The particular pyroxyline recommended by Wallace is not obtainable in England, but Hopkins and Williams's high temperature pyroxyline gives excellent results, as does also Mawson and Swan's collodion when precipitated in this way. It is not necessary to use amyl acetate before the precipitation, the ordinary solvent of equal parts of alcohol and ether proving quite satisfactory, and the precipitation is instantaneous. When thoroughly dry the precipitated pyroxyline is again dissolved in the above proportions, and the collodion carefully filtered through paper. This, which is trouble some unless pressure is used, may be avoided by allowing the collodion to stand three or four days, when the whole of the impurities settle to the bottom of the bottle.
The grating should be carefully levelled in a drying cupboard, and this is absolutely essential to prevent the occurrence of dust ; next, it is carefully dusted with a soft camel-hair brush, and the solution flowed over the surface. The exact quantity is a matter of experiment, too thin a film being difficult to handle whilst too thick a film gives a matt surface. About 1-5 ccs. is the right quantity to use for a 2-in. grating. The best method of applying the solution is with a fine pipette, which will hold just the necessary quantity. This enables one to distri bute the solution over the surface without touching the grating itself, and the solution can be easily led to any part, or an air bubble brought to the edge and broken.
The coated grating should be left in the cup board for at least twelve hours, and longer is preferable, even for three or four days ; the longer it is left the easier it is to handle. To
strip the cast, the grating should be placed in a dish of distilled water, when the edges will soon begin to show shadow bands. As soon as these are observed the grating should be taken from the dish, and any adherent water removed with a soft rag. Slight pressure with the thumb nail along one edge will cause the cast to spring from the metal, and it should then be grasped by a pair of wide-jawed forceps, as used in micro scopy, and pulled off with a firm but even motion in a direction parallel to the lines of the ruling. The edges of the cast should be trimmed off, and it should then be lowered on to a piece of carefully cleaned and polished plate glass, which should be immersed in distilled water and lifted out with a small pool of water on its surface. One edge of the cast should be lowered on to the glass first, and then the rest gradually lowered so that it pushes the water in front of it without the occurrence of air bubbles. As soon as it is in position one edge should be clipped by a strong metal clip, and a piece of soft velvet rubber passed across it in the direction of the rulings ; when contact is obtained everywhere the edges should be cemented down with some of the collodion as used for the cast. This can be applied with a very fine camel-hair brush. The cast may be cemented face up or face down, but with the latter there is less chance of the rulings being damaged. When the cast has been cemented it may be dried by heat, gentle at first, but gradually increased to 167° P. (75° C.). The chief cause of failure is dust particles between the glass and the replica. If the replica is mounted face down, another piece of glass may be cemented to the back.
There is some contraction of the cast in drying, but this is small. Wallace found that with a grating of 28.867 mm. width and 16,397 lines the shrinkage was 0•176 mm. on the entire width, or about six lines more to every one thousand. The shrinkage can, of course, be easily deter mined by accurate measurement of the original grating and the cast ; thus, in the case referred to above, the original width = 28.867 mm., width of replica = 28.691 mm. ; therefore with a total of 16,397 lines in the original, 568 lines = i mm. On the replica obviously 16,397 28.691 = 572 per mm. This shrinkage simply causes greater dispersion of the spectrum.
When examined in a quartz spectrograph these replicas were found to transmit up to x 2,613 in the ultra-violet, but obviously their glass support absorbs up to about A 3,400. Prof. R. W. Wood has suggested the use of thin mica sheets, and naturally quartz could be used for the support. This process of taking casts from a grating in no way damages it, it being, in fact, an excellent method of cleaning a grating.