February 1, 1894.] 



KNOWLEDGE. 



39 



Dr. Matthiessen predicted the time of perihelion passage 

 in 1800 to be May 9th, and gave the probable error of his 

 determination as 'less than four days. From this it would 

 appear that the orbit of the comet has been computed with 

 considerable precision, though the available observations 

 of it in 1881 only extended over the fifty days from October 

 5th to November 24th. The exact periodic time was given 

 by Hartwig as 8-8335 years, by Plummer as 8-8578 years, 

 and by Matthiessen as 8-6874 years. Chandler made it 

 less, while Block at Odessa made it more. Of the several 

 values, Matthiessen's is probably the most accurate, as he 

 utilized an observation at Strasburg, on November 24th, 

 not known to the other computers. If we adopt his period 

 of 8-6874 years ( = 8 years and 251 days), as correctly 

 representing the mean period of the comet, then returns 

 may be expected in January, 1899, September, 1907, June, 

 1916, February, 1925, October, 198H, &c. 



Including a few comets omitted by Mr. Lynn as too 

 uncertain, and adopting the periodic times given in 

 AstmnoDii/ awl Astni-Phi/'fics for November, 1893, I have 

 compiled' the following table of probable cometary returns 

 up to the end of 1900 : — 



18114, Aiiril, Temxjel (1ST3 11.) 



ISilo, Febi-j'.. Eucke. 



June, Barnard (1S84 II.) 



Jiily, Brorsen. 



IS'.lti, Marcli, Faye. 



October, Brooks (1889 V.) 



IsilT, March, Spitaier (1890 II.) 



May, D'Ai-rest. 



May, Tempel-Svvift. 



August, Brooks (1886 IV.) 



1.S9S, April, Pons-Wimiecke. 



May, Eneke. 



I have given the months of perihelion passage, but of 

 course they are very doubtful in many cases. Their re- 

 tention, however, serves to indicate the season. Wherever 

 it seemed necessary to distinguish the comet by adding the 

 year and number to the name, I have done so. The list 

 shows that returns of periodical comets will be remarkably 

 frequent in the last three years of the present century. For 

 Tempel's comet (1866 I.) I have adopted Dr. Oppolzer's 

 period of 83-176-years, but the mean period is probably a 

 little longer than that, and corresponds with the 33^ years 

 of the Leonid meteors. Yours faithfully, 



Bristol, -January 8th, 189i. W. F. Denning. 



1.S98, June, Swift 1 1889 VI.) 



-June. Wolf. 



Septr., Tempel (1867 11.) 



1899, January, Denning (1881 V.) 

 March, Tempel (18615 I.) 

 April, ]5arnard (1892 V.) 

 May, Tuttle (1858 1.) 

 May, Holmes (189-2 III.) 

 July, Tempel (187:) II.) 



1900. Fabry., Finlay. 

 October, Barnard (1884 II.) 



Re LARGE TELESCOPES. 

 To the Editor of Kno-wledge. 



1)e.\r Sir, — In your interesting paper upon telescopes, in 

 this month's Knowledge, you state that Lord Rosse's Oft. 

 mirror remains " the greatest light-grasping instrument 

 which has been turned to the stars." 



Surely you forgot Dr. Common's 5ft. silver on glass 

 reHector. There can be no doubt of its superiority as a 

 light-collector over a Oft. metal reflector. The highest 

 estimate I have found for Lord Eosse's metal is that it 

 returns 67 per cent, of the incident light. The second 

 reflection at the flat will be -67 x -67, or -45 about. 



A silver on glass in good condition reflects 91 per cent., 

 and its effective return after second reflection will be 

 •91x-91 = -83 nearly. We compare the light-gathering 

 power of the two instruments by multiplying the square 

 of the diameters of -45 and -83 respectively then, which 

 comes out as 161-0 for Lord Rosse's, and 207 for Dr. 

 Common's, or the latter collects more than one-fourth 

 more light. 



It appears to me that neither large refractors nor large 

 reflectors are quite satisfactory, but there is no doubt 



the American refractors have much better atmospheric 

 conditions than our English reflectors. 



January 5th, 1894. Edwin Holmes. 



[Mr. Holmes is, no doubt, correct when he takes into 

 account the loss of light on a second reflection at the 

 surface of a flat. But the instruments may be used with 

 a totally reflecting prism in place of a flat, or a photo- 

 graphic plate may be exposed in the principal focus. 



When only the one reflection at the speculum is taken 

 into account. Lord Rosse's instrument has the greater 

 light-grasping power. Adopting Mr. Holmes' estimates 

 for the percentage of light lost by reflection at perpendicular 

 incidence from speculum metal and from silver, the ratio 

 of light-grasping power would be as 36 x -67 to 25 x "91, 

 or as 24-12 to 22-75. 



In comparing the loss of light when a flat is used, it 



should be remembered that less light will be lost when 



the angle of reflection is 45° than when the light is 



reflected at nearly perpendicular incidence, as from a 



speculum, and that a silver on glass mirror tarnishes 



much more rapidly than a speculum metal reflector ; so 



that in ordinary use a silver on glass reflector stands at 



a greater disadvantage, as compared with a newly-polished 



instrument, than a speculum metal reflector. — A. C. 



Ranyard.] 



— I • i — 



To the Editor of Knowledge. 



Dear Sir, — I beg to draw attention to a curious optical 

 phenomenon, which seems to me worthy of more attention 

 than it appears to have received. 



When the eyes are fixed on a light at some distance, 

 and partly closed, three pencils of rays, in the form of the 

 letter Y, are usually seen apparently proceeding from it. 

 One of the upper pencils disappears when the corresponding 

 eye is closed, together with half of the lower pencil. 

 Each pencil is seen to be made up of a varying number of 

 separate rays. 



I have found by experiment that the apparent arrange- 

 ment of these rays varies remarkably, according to the 

 structure of the eye of the observer. When the eye is 

 normal the rays proceed directly from the luminous point 

 of the flame ; when myopic the rays cross each other and 

 separate a short distance from the luminous point ; when 

 presbyopic the rays meet (or would if produced) beyond 

 the luminous point. A concave or convex lens will enable 

 anyone with normal eyes to see the appearances peculiar 

 to presbyopic or myopic eyes respectively. 



It is a curious fact that the arrangement of these rays 

 in a particular eye is in fact a visible representation of the 

 manner in which the actual rays reflected from outer 

 objects traverse that eye, if we suppose the retina to 

 occupy the place of the luminous point. 



I suggest the following as a possible explanation : — The 

 rays emitted from a luminous body, besides throwing an 

 image of the body on the retina, have also the power of 

 giving the sensation of an addition image, which takes 

 the form of two pencils of rays proceeding from the 

 luminous point. In a myopic eye the actual image as 

 seen is badly defined, owing to well-known causes, and 

 instead of a single image, a number of partly superimposed 

 similar images are seen, each with its complemental 

 pencils of rays. These rays cross each other and give 

 the observed appearance. If a concave lens is now inter- 

 posed between such an eye and the luminous body, and 

 moved towards the eye so as to gradually correct its 

 refraction, the images will approach, and form at length a 

 definite image, giving two pencils of rays arranged exactly 

 as in the normal eye. If the lens is brought still closer 



