October 20, 1904] 



NA TURE 



615 



of the viewing' lens. The white area in the centre was of 

 good quality, though not quite so bright as in the photo- 

 graphic gratings, for the reason that the three sets of 

 rulings were rather more than the glass surface would 

 take without breaking down between the lines. Photo- 

 graphic copies of this multiple ruling have been made, and 

 will probably prove useful in demonstrating colour 

 synthesis. 



The appearance of two overlapping rulings under the 

 microscope is shown in Fig. ib. 



It would appear at first sight as if a ruling of this de- 

 scription would be incapable of giving distinct spectra, and 

 we should certainly not expect it to give merely the super- 

 posed spectra of the two gratings. 



.^s a matter of fact, secondary spectra are produced, 

 though they are usually so faint that they give no trouble. 

 In some cases, however, owing to some peculiarity of the 

 form of the groove, the photographic copies when super- 

 posed do not give the expected colour. For example, in one 

 exceptionally pronounced case, the superposition of the red 

 and violet gratings gave, instead of purple, a brilliant 

 yellow-green. 



The origin of the secondary spectra can be seen in the 

 following way. If the red and violet gratings are super- 

 posed with the lines mutually perpendicular, and a lamp 

 is viewed through the combination, the spectra appear as in 



Fig. 2, the secondary spectra being usually much fainter 

 than the principal ones. If, now, one of the gratings is 

 slowlv rotated through a right angle, the spectra will 

 gradually wheel around into a straight line, and the 

 secondary spectra will be found to fall between the principal 

 spectra. In the particular case alluded to above, the 

 secondary spectra for some reason or other were brighter 

 than the principal, and it was found that the yellow-green 

 of one of them fell at the point where the red and violet 

 of two of the principal spectra coincided. This accounted 

 for the abnormal colour which appeared in this case. It 

 is very seldom, however, that these abnormal colours appear 

 in the pictures. 



.\ set of gratings for the production of colour photographs 

 was ruled on this machine, and results far superior to any 

 that had ever been obtained before were immediately secured. 

 A few of the pictures were of such excellence as to compare 

 favourably with the results obtained in the Kromshop. The 

 method of preparing the pictures was essentially as I have 

 described in previous papers. 



The diffraction process has also been successfully com- 

 bined with the Joly process. To accomplish this it was 

 necessary to rule the three sets of lines in bands correspond- 

 ing to the width of the red, green, and blue lines of the 

 Joly screen. Calculation showed that if 12 lines were ruled 



NO. 1S25, VOL. 70] 



with the 5-tooth cam, 10 with the li-tooth, and 9 with the 

 5-tooth, the spaces would be about right. 



Various schemes for making a ruling of this description 

 were considered, but no satisfactory automatic device 

 appeared to be possible, since the period of the bands on the 

 Joly screen could not be exactly duplicated by any combin- 

 ation of cams. 



The following simple device was finally hit upon. The 

 engine was equipped with the 7-tooth cam, and a small 

 stepped piece of brass mounted under the lever which 

 operated the pawl, which, by preventing the complete de- 

 scent of the lever, caused an advance of only 6 teeth or 

 5 teeth according to its position. The Joly screen was 

 mounted on the table of the engine under a microscope, and 

 the transit of the coloured lines across the X hair in the 

 eye-piece observed, the rate of the ruling being changed at 

 the proper moment by sliding the stepped piece of brass 

 into the proper position, which was done by means of a 

 short brass rod. The ruling of the grating occupied twelve 

 hours, during which time I was obliged to sit with my 

 eye constantly at the microscope, for the change of rate 

 occurred about every half minute. Two very satisfactory 

 gratings were prepared in this way, one corresponding to 

 the Joly screen and the other to one of the screens ruled on 

 the machine of the McDonough Co., of Chicago. 



These gratings when placed in the viewing apparatus 

 appeared w-hite and very brilliant, and were easily duplicated 

 b\ photography. 



They were used in the following way : — 



A positive on glass, made from a Joly negative, w-as 

 flowed with a thin solution of gelatin sensitised with 



Fig. 



bichromate of potash, and allowed to dry. The triple ruled 

 grating was then placed with its ruled surface in contact 

 with the sensitive film, and held before the lens of the 

 viewing apparatus. The appearance of the picture was now 

 precisely similar to the appearance when a Joly colour 

 screen was used, and the lines could be brought into register 

 at once, when the picture appeared in its natural colours. 

 A ten-second exposure to the light of the arc recorded the 

 grating lines on the film, and the plate was then dipped into 

 warm water and dried. The colours of the picture prepared 

 in this way were fully equal, if not superior, to those 

 obtained with the Joly viewing screen. There is the added 

 advantage that the colour lines and picture lines are on one 

 and the same film, consequently there is no liability of the 

 lines to get out of register. Moreover, the picture can be 

 duplicated by contact printing on glass sensitised with 

 chrom-gelatin. These prints are, of course, quite trans- 

 parent until they are placed in the viewing apparatus, when 

 the coloured picture at once appears. 



By this means the trichromatic screen, w^hich in the Joly 

 process must be mounted with every picture, is done away 

 with, and a picture obtained which can be easily duplicated. 

 To offset the advantages gained we have, however, the dis- 

 advantage that the pictures require the viewing apparatus, 

 and show the same obtrusive lines as the original Joly 

 pictures. The triple diffraction screen could easily be ruled 

 with its colour elements much closer together, but we should 

 gain nothing in this way until some method of taking a 

 negative with narrower colour elements is devised. 



The pictures made by the old method with three separate 



