i;,o 



NATURE 



\yune 13, 1872 



the captures of American species take place in Ireland in autumn 

 and early winter, and in all cases are species belonging to the 

 northern portion of America which migrate southward at the 

 close of the breeding season. At that time the prevailing winds 

 are from the west, and tlie birds in their flight become confused, 

 and are carried across by the winds, taking an occasional rest on 

 passing vessels. 



Mr. J. ConniN', of Durban, forwards to the Nalal Colonist \.]m 

 following account of a "sea-serpent" seen by him: — "During 

 my late passage from London, I saw no less than three sea- 

 serpents, but an account of the last will suffice. On 30th De- 

 cember last, on board the Silvery IVai'f, in lat. about 35' o' ,S., 

 and long. 33' 30" E. , at 6.20 P.M. solar time, an enormous ser- 

 pent passing nearly across our bows co.mpelled the alteration of 

 our course. He was at least one thousand yards long, of which 

 about one-third appeared on the surface of the water at every 

 stroke of his enormous fan-shaped tail, with which he propelled 

 himself, raising it high above the waves, and arching his back 

 like a land-snake or a caterpillar. In shape and proportion he 

 much resembled the cobra, being marked by the same knotty and 

 swollen protuberance at the back of the head on the neck. Tlie 

 latter was the thickest part of the serpent. His head was like 

 a bull's in shape, his eyes large and glowing, his ears had circular 

 tips and were level with his eyes, and his head was surmounted 

 by a horny crest, which he erected and depressed at pleasure. 

 He swam with great rapidity and lashed the sea into a foam, like 

 breakers dashing over jagged rocks. The sun shone brightly 

 upon him ; and with a good glass I saw his overlapping scales 

 open and shut with every arch of his sinuous back coloured like 

 the rainbow," 



T 



ABNORMAL DISPERSION OF BODIES WITH 

 S UREA CE-COLO URS, '■> 



HESE memoirs relate to the abnormal phenomena of dis- 

 persion produced by certain bodies which reflect the rays 

 of some colours in the manner of transparent media, and those 

 of other colours in the manner of metals. 



In transparent bodies the velocity of light is less than 'in a 

 vacuum, and the inde.\ of refraction increases as the length of 

 undulation diminishes, whereas, in the case of metallic bodies, 

 theory leads to essentially different results. From the researches 

 of Jamin on the elliptic polarisation of light reflected by 

 raet.ils, and from the more recent experiments of Quincke, it 

 appears that the formula; of Cauchy, Beer, and Eisenlohr give 

 for certain metals, as gold and silver, an index of refraction less 

 than unity, whence it follows that in these metals the velocity of 

 light is greater than in the vacuum. Cauchy's theorem, verified 

 by the experiments of J.\min, likewise shows that the index of 

 refraction in metals varies with the angle of incidence, and that 

 in most metals dispersion takes place in a manner opposite to 

 that which occurs in transparent bodies, that is to say, that the 

 rays of sliortest wave-length are less refracted than those wliose 

 undul.ttions are longer. 



The only physicist who has hitherto attempted a direct verifi- 

 cation of these theoretical results is Quincke ; but his experi- 

 ments h.ive yielded contradictory results according to tlie 

 method employed, and have therefore afforded no decisive in- 

 formation as to the indices of refraction and dispersion of metals. 

 Kundt, in studying the properties of a p.articular class of bodies 

 which approach the metals in their optical properties, has suc- 

 ceeded, if not in determining the absolute values of their refrac- 

 tive indices, at least in observing certain anomalous phenomena 

 of dispersion, which appear to indicate the direct road towards 

 a new verification of the theoretical formula;. 



The bodies in question are media intermediate between trans- 

 parent bodies and metals, inasmuch as they behave like metals 

 towards rajs of ceriain colours, and like transparent bodies 



"* By A. Kundt (Pogg. Ann , cxlii. 177 ; Journal de Physique, 1872, p. 38) 

 nd J.' L. .Syret (Pogg. Ann., c.tliii. 325 ; Journal de Physique, 1872, p. 45). 



towards rays of other colours ; they aUo reflect light with a 

 certain degree of metallic lustre. These substances may be in- 

 cluded unrler the denomination of Iwdia with surface-colours. 

 Most (but not all) of them are strongly tinctorial, and exhibit, 

 both in solution and in small fragments, a beautiful coloured 

 transparency ; such are most of the aniline-dyes, indigo, cartha- 

 mine, potassium-permanganate, &c. 



The optical properties of these bodies have been studied Ijy 

 Brewster, Haidinger, Stokes, and several others, tlie chief re- 

 sult of whose researches is the law, first enumerated by Hai- 

 dinger, that the light transmitted by these media is exactly or 

 nearly complementary to the light reflected from their surfaces, 

 and, therefore, to the colour of the surface. Moreover, Dale and 

 Baden Powell observed that indigo and Prussian blue, in re- 

 flecting light, polarise it elliptically, like the metals; and von 

 der Willigen determined the refleciiim-constants of pale indigo, 

 and found that the principal angle of incidence diminishes from 

 the line B to E, and then increases from E to G. 



The author's views respecting these bodies are based on the 

 hypothesis that they exhibit the most general case of dispersion, 

 that, namely, in which the index of refraction may not only in- 

 crease or diminish when the wave-length in air diminishes, but 

 may even become less than unity. 



The examination of the light reflected by these bodies is alone 

 sufficient to lead to this hypothesis. For when a medium is 

 transparent for a particular ray, the intensity of the reflected 



light is expressed by the formula ( V the value of 



which increases or diminishes with the increase of n, according 

 as II is greater or less than unity. In most transparent bodies 

 the variation of n for the different colours, — that is to say, 

 the dispersion — is so small, that the whole of the reflected light 

 exhibits lire same colour as the incident light, so that, if the in- 

 cident light is white, the reflected light is white also. Applying 

 this formula to bodies with surface-colours, it would follow that 

 the rays which form the superficial colour, that is to say, those 

 which are reflected in very great proportion, should have, 

 relatively to the others, a very large or very small index ol 

 refraction. 



As the rays which form the surface-colour may have any 

 wave-length whatever, it would follow from the formula which 

 gives tlie intensity, that the dispersion of bodies with surface- 

 colours should be, or at least might be, abnormal, in a manner 

 whicli is perfectly arbitrary. It might even happen that a 

 portion of tlie transmitted rays, if made to pass witli a sufficient 

 intensity through a prism of the substance, would be refracted on 

 one side of tire direction of the incident rays, and the rest on the 

 other side. In fact, for rays which are reflected in large propor- 

 tion, that is to say, rays for which the bodies under consideration 

 act like metals, the preceding formula is not applicable, and it is 

 necessary to employ the formulce given by Cauchy, in his theory 

 of metallic reflection. 



These formulae give, for the normal incidence : 



(*--:) 



where ij/ is determined by the relation 



cotan ij( = cos 6 sin 2 tan -^ e, 

 c and 6 being two constants to be determined by experiment, 

 according to the relations — 



6 cos 6 = « ; 8 sin € = 7, 



// being the index of refraction, and 7 the coefficient of extinction 

 under the normal incidence. 



According to these formula; it is not absolutely necessary, as 

 remarked by Cauchy, that the index of refraction of metals 

 should be very large. Nevertheless it appears from the experi- 

 ments of Jamin, and those more recently made by Quincke, that 

 the constants of elliptic polarisation have values which indicate 

 for most metals a very large refractive index, and for silver and 

 gold an index less than unity. 



Cauchy's formula:, applied to bodies with surface-colours — ad- 

 mitting that the- elliptic polarisation produced by these bodies is 

 analogous to that produced by the metals — renders pronab'e the 

 existence of refractive indices cither very large or less than 

 unity : hence the probability of abnormal dispersion resulting 

 from transmission through ihese bod es. 



When one of these surface-coloured bodies is dissolved, it 

 preserves in solution its abnormal properties relatively to dis- 

 persion. This abnormal dispersion will be combined vnth the 



