6i4 



NATURE 



[October 20, 1904 



Argonian (7 Argus). — H, p H. 

 Ahiilamian (e Orionis). — H, He, A4649, Si IV. 



Crucian (a Crucis). — H, He, 



Ast, O, N, C. 

 Tanrian (f Tauri).— H, He, 



p Mg, Ast. 

 Rigdian O Orionis).— H, 



p Ca, p Mg, He, Si II. 

 Cygnian (« Cygni).-H, p Ca. 



p Mg, p Fe, Si II., p Til 



pCr. 

 Polarian (a Urs. Min.).— 



p Ca, p Ti, H, p Mg, p Fe, 



Ca, Fe, Mn, Si I. 

 Aldebarian (a Tauii).— p Ca, 



Fe, Ca, Mn, p Sr, H, Si I. 

 Antarian (a Scorpionis). — 



Flulings of manganese,^ and 



many metallic lines. 



INebuL-e.] 



1 Many of th's lutings have sine 



[Fowler, Roy. Soc. Proc, vol. Ixj 



most strongly developed in the less i 



are not seen in the spectrum of Betelgi 



Acker. 



(» 



Eridani).— 

 'ersei).— H, 



Algolian W 



p Mg, p Ca, Me, Si 11. 



Markabian (a Pegasi).— H, 

 p Ca, p Mg, Si II. 



Sirjan (a Canis Maj.).— H, 

 pCa, p Mg, p Fe, Si II. 



Procymiian (a. Canis Min.). 

 —Same as Polarian. 



^hcian (19 Piscium).^Flu- 

 tings of carbon and many 

 metallic lines. 



been 



[Dark S 



lown to be due to tita 



ii. p. 219 (1904)]. The fluting! 



frangible parts of the spectrum, 



reproduced in Fig. 



E.\amples illustrating most of ttie groups are given in 

 Figs. 4 and 5, from negatives tatcen by Sir Norman Lockyer 

 and his assistants at the Solar Physics Observatory. These 



are to appear coloured are represented by diffraction gratings 

 of various spacings. A grating ruled on glass, when com- 

 bined with a convex lens and directed towards a lamp flame 

 or other source of light, forms diffraction spectra in the 

 focal plane of the lens. If the pupil of the eye is brought 

 into the red portion of one of these spectra, we perceive 

 the entire surface of the grating illuminated in red light, 

 since every portion sends red light, and red light only, into 

 the eye. If a second grating with closer ruling is substi- 

 tuted for the first, the eye remaining fixed in position, the 

 spectra will occupy different positions, and if the pupil of 

 the eye occupies, say, the green region of one of them, this 

 grating will appear green. If the two gratings are placed 

 side by side, and overlapping one another, the one will 

 appear red. the other green, while the overlapping region, 

 since it sends both red and green light to the eye, appears 

 yellow (secondary yellow). If a third grating of still finer 

 spacing is now placed before the lens, partly overlapping 

 the other two, it will appear illuminated in blue-violet light, 

 and the overlapping portions will be coloured purple, white, 

 and bluish-green. 



We may in this way obtain a large variety of colour 

 with only three rulings, and since the intensity of the light 

 depends on the distinctness with which the lines are ruled 

 or photographed, light and shadow can be obtained solely 



; of decreasing temperature 

 Majoris (Sir 



-(1)7 Orionis (Cru. 

 n) ; (5) a Canis Mil 



bring out very clearly the gradual simplification of the 

 spectrum in the first series as the temperature rises, and 

 the increasing complexity in the second series as the 

 temperature falls. On the dissociation hypothesis, we have 

 first to deal chiefly with relatively cool metallic vapours, 

 which, as the temperature rises, are brought by dissociation 

 to the proto-metallic stage, and finally to the gaseous con- 

 dition represented by hydrogen and helium ; then, through 

 subsequent cooling, association begins and produces some- 

 what similar changes in inverse order. 



A. Fowler. 



RECENT IMPROVEMENTS IN THE DIFFRAC- 

 TION PROCESS OF COLOUR-PHOTO- 

 GRAPHY.^ 



T^HE fundamental principles of the diffraction process of 

 colour-photography will be found in my earlier papers 



on the subject." In brief, the method consists in preparing 



by photographic means a picture in which the areas which 



1 Paper read before Section A of the British Association at the Cambridge 

 meeting by Prof. R. W. Wood. 



- Wood, " Application of the Difl^raction Grating to Colour-photography" 

 (P/iiV. jlfag., April, 1S99); "Diffraction Process of Colour-photography," 



(Nat I 



ol. Ix. p. 199, 1899). 



NO. 1825, VOL. 70] 



n); (2)fi Persei (Algoli: 

 is (Procyonian) ; (6) a. Ei 



by the presence of the diffracting lines. The portions of the 

 plate on which they are absent send no light to the eye, and 

 appear black. 



A full description of the method by which photographs 

 showing the colours of the original object were prepared 

 will be found in the papers above referred to. 



The earlier experiments were made with very imperfect 

 gratings, the periodic errors of which caused the pictures 

 to show vertical bands of colour. During the past winter 

 I have ruled gratings of various description on one of the 

 Rowland engines, and continued the e.xperiments of five 

 years ago. 



This machine was designed to rule 14,438 lines to the 

 inch, but by employing larger cams, which cause the pawl 

 to skip a specified number of teeth, it may be made to rule 

 at the rate of 7219, 4812, 3609, and so on. Calculations 

 showed that gratings ruled on this machine with cams 

 which advanced the toothed rim of the large wheel five, six, 

 and seven teeth respectively would be suitable, that is, 

 would have the relative spacings necessary to produce white 

 when they were superposed. 



To illustrate the principle of the colour synthesis, a glass 

 plate was ruled with the three spacings, the ruled squares 

 overlapping as shown in Fig. la. The areas appeared 

 coloured as indicated when the plate was placed in front 



