November 23, 1905] 



NA TURE 



79 



the gelatin, and then a slow diffusion of the remaining salt 

 lakes place downwards, and this might give rise to the idea 

 that the thing was really growing. 



W. A. Douglas Rl'dge. 

 Woodbridge School, Suffolk. 



The Spectrum of the Positive Rays (Canal.Strahlen . 



In former publications (Ann. d. Phys., vol. xiv., p. 524, 

 1904; vol. xvi., p. 490, 1905; " Die Elektrizitat in Gasen," 

 Leipzig, 1902, pp. 44b, 457) I have expressed the opinion 

 that the carrier of the line spectrum of a chemical element 

 is the positive atom-ion, while, on the other hand, the band 

 spectrum is due to the re-combination of positive atom-ions 

 with negative electrons. From this it would follow that 

 the particles of the positive rays (being positive atom-ions) 

 should emit the line spectrum of the gas in which they 

 are produced. Moreover, since these particles possess a 

 considerable velocity, their spectrum lines, observed in the 

 direction of the rays, should have a position and breadth 

 differing from the position and breadth when the 

 lines are observed in a direction normal to the rays. 

 Again, since the positive rays ionise the gas traversed by 

 them, re-combination of positive and negative ions must 

 take place in their path ; the gas traversed by the positive 

 rays must therefore emit the band spectrum, which will 

 be superposed on the line spectrum due to the positive rays 

 themselves. The lines of the band spectrum must have 

 the same position whether they are observed in the direc- 

 tion of the positiv.' rays or in a direction at right angles 

 to them, inasmuch as their carriers do not possess the 

 velocity of the positive rays. 



I now state briefly the results of an experimental spectro- 

 graphic examination of the light emitted by a gas traversed 

 by positive rays. (1) Nitrogen shows simultaneously the 

 band and the line spectrum, hydrogen shows simultaneously 

 the series spectrum (Ha, H/3, . .) and the many lines 

 spectrum. (2) The line spectrum emitted normally to the 

 positive rays and the line spectrum emitted in the direction 

 if the rays are different; the former shows in hydrogen 

 sharp lines of the known wave-length, while the latter 

 shows these " stationary " lines, and besides them, on their 

 ultra-violet side, new widened lines (" displaced " lines). 

 (3) This displacement is greater when the velocity of the 

 positive rays is greater. (4) The lines of the band spectrum 

 (many lines spectrum) have the same position and breadth 

 whether they are observed in the direction of the positive 

 rays or in a direction at right angles to them. A full 

 account of the investigation will be published shortly. 



Gdttingen, November 3. J. Stark. 



Replicas of Diffraction Gratings. 



Kindly allow me to correct a statement contained in 

 your notice of Mr. R. J. Wallace's replicas of diffraction 

 gratings in your issue of November 2 (vol. lxxiii. p. 21). 



It is there stated, as also in the Astrophysical Journal 

 from which the extract was taken, that I first flood the 

 grating with oil in my method of producing replicas. This 

 1 may say I have never done except when making experi- 

 ments, my procedure being exactly the same as Mr. 

 Wallace's, viz. to flood the grating direct with the clarified 

 celluloid solution, dry it in much the same way, but using 

 special precautions to ensure perfectly even drying, stripping 

 and mounting in a similar manner to Mr. Wallace, but 

 leaving out the gelatin coating, which in my opinion is 

 quite unnecessary. 



[■beg to enclose you one of Mr. Wallace's first quality 

 (average) replicas, kindly sent to me by him in exchange 

 for one of my own, as also a couple of mine for comparison. 

 The great difference to be noted in their surfaces and per- 

 formance is due to the peculiarities in the surfaces of the 

 original gratings, one of my own replicas having a bright- 

 ness in the first spectrum on one side of at least four times 

 that of the other, and twice that of Mr. Wallace's replica. 



The grating from which the very bright replica is taken 

 is a " Rowland " of 14,438 lines to the inch, and was 

 formerly the property of the late Dr. Common. The 

 original of the other is a very beautiful specimen of recent 

 work on the Rowland engine, 15,038 lines to the inch. 

 Now, whilst the latter when mounted on parallel plane glass 



NO. 1882, VOL. 73] 



gives comparatively feeble spectra, when mounted on prisms 

 lor direct-vision purposes, and tilting the prism to the 

 angle required for the minimum deviation for the diffrac- 

 tion spectrum, first order, the brightness approai hes that 

 from the " Common " grating, whilst its much greater 

 freedom from scattered light renders it very suitable for 

 prominence and similar work, the dispersion being about 

 equal to five 6o° flint glass prisms in the centre of the 

 spectrum, and decidedly greater at the red end. 



This increase of brightness is, of course, attributable to 

 the form of the grooves, less interference being produced 

 under the latter condition, and this notwithstanding the 

 increase in dispersion. 



It may be of interest to some to know that I have 

 succeeded in mounting these grating films on a perfectly 

 flattened ring of glass, so that, by avoiding the use of 

 glass as a base, light of very short wave-length can be 

 examined by this means, either in the one case by trans- 

 mission to about X 2600 or by reflection to as low as 

 \ 1850, and possibly lower. (The discovery of this re- 

 flective property for ultra-violet light was made by Mr. 

 Morris-Airev, of the Victoria University, last year.) In 

 order to examine by reflection either a partial vacuum is 

 created behind the film when mounted on a glass ring or 

 the film is mounted on a concave surface, which, although 

 not giving the lines of the grating their true form, gives 

 very fair resolution. 



1 have also succeeded in making concave replicas prac- 

 ticall) as perfect as plane ones, by rotating the grating 

 during the drying process at such a rate that the para- 

 boloid'al curvature of the solution was practically the same 

 as that of the grating. Anyway, the difference is so slight 

 thai when dry no rings can be seen on examining it by 

 monochromatic light before the film is stripped from the 

 grating. The difficulty of silvering these replicas satis- 

 factorily has, however, prevented further progress, for the 

 present at least. 



In justice to Mr. Wallace I ought to say that, in a reply 

 1.. .1 letter from me, he states he obtained his information 

 from a patent I once took out in connection with the 

 application of these grating replicas to colour photography ; 

 but the method there described is not the one I have 

 adopted in making my replicas. 



Thomas Thorp. 



Whitefield, near Manchester, November 6. 



The article referred to by Mr. Thorp was, as mentioned 

 in its first two lines, simply a resume of Mr. Wallace's 

 article in the Astrophysical Journal ; and the statement 

 corrected by Mr. Thorp was taken from that journal. It 

 may be said, however, that the three gratings sent by 

 Mr! Thorp have been compared, under similar conditions, 

 and the results are in full accordance with the descriptions 

 given above. 



The second " Thorp " replica (15,120 lines) is a beautiful 

 specimen in appearance, having none of the mottling which 

 appears on the 14.438-line copy, but its first-order spectra 

 are excelled in brightness by the first-order spectrum on 

 one side of the latter, which is a similar replica to those 

 that have produced such remarkably good results in eclipse 

 and other observations when attached to just an ordinary 

 hand-camera or a simple modification thereof. 



The Writer or the Notice. 



Aurora of November 15. 



An extraordinary aurora was seen here on November 15 

 at 6 p.m., in appearance something like a " stormy " 

 sunset. The lower part of the aurora was illumined with 

 a bluish-green light, and had an altitude at the centre of 

 about 10°. Above this, extending for a further ten or 

 fifteen degrees, the sky was brilliantly illuminated with 

 streamers of a rosy red colour. I did not wait for its 

 disappearance, but at 9.30 p.m. red streamers were visible 

 in ih- N.N.W. 



I looked for the Leonids this morning continuously from 

 1.20 to 2.35 from the vantage ground of the Dartmoor 

 hills (1500 feet). I saw only one Leonid at 2.30, but 

 the moon would prevent small ones being seen. The night 

 was perfectly clear. Rowland A. Earp. 



Buckfastleigh, S. Devon, November 16. 



