Theory of Diffraction-gratings. 89 



but, on the other hand, it will not do to leave it long enough to 

 become sensibly moist. Something will probably depend on the 

 particular sample of collodio-chloride. 



The exposure required is about five or seven minutes to the 

 autumn suu. On a hazy clay something more may be required ; 

 but if there are many clouds about, the experimenter, unless na- 

 turally of a good temper, will do well to postpone his operations. 



On removal from the frame, the plates may be placed in a dish 

 of water until it is convenient to finish them. They are fixed, 

 without any toning, in a dilute solution of hyposulphite of soda, 

 such as is used for paper prints, and then carefully washed. 

 The most effective washing is a combination of rinsing and soak- 

 ing. My practice is to rinse the plates under the tap for half a 

 minute in order to remove the greater part of the hyposulphite 

 of soda, and then to allow them to soak for an hour or two in 

 water changed two or three times. After a final rinsing the 

 plates may be set up to dry. 



Gratings finished in this way give excellent definition, but the 

 spectra are rather deficient in brilliancy. This defect is of less 

 importance than might be supposed; for in order to see the 

 finer fixed lines, sunlight is in any case indispensable, and with 

 sunlight there is usually illumination to spare. Nevertheless, as 

 gratings are likely to be largely used for the purpose of popular 

 illustration under circumstances where artificial light must be 

 employed, I am glad to be in a position to recommend a simple 

 mode of treatment by means of which the brilliancy of the spectra 

 may be materially enhanced. For this purpose it is only neces- 

 sary to treat the fixed and washed impression with a solution of 

 corrosive sublimate. When the whitening effect is complete, 

 the plate must be again washed and then set up to dry. Con- 

 sidered as a photographic transparency, the grating is reduced 

 rather than heightened in intensity by this process. The cause 

 of the improvement of the spectra will be touched upon presently. 

 These mercury-treated gratings cannot be varnished without 

 sacrificing most of the advantage of the method. I have occa- 

 sionally applied the same treatment successfully to tannin plates. 



When not in use, the finished gratings should be kept in a 

 dry place and protected from dust and other atmospheric de- 

 posit. For this purpose they may be put away wrapped in paper. 

 For a short time there is no objection to leaving them standing, 

 face inwards, against a wall ; but a better plan is to place them, 

 face downwards, on a flat and thoroughly clean piece of plate 

 glass. 



The originals from which 1 have hitherto taken copies are 

 three in number. Two are by Nobert, and contain respectively 

 3000 and 6000 lines, in each case covering a square inch (Paris) 



