Sept. 15, 1887] 



NATURE 



461 



Cocoa-nut Pearls. 



Referring to the letter of Dr. Sydney J. Ilickson, published 

 in your paper of June 16 last (p. 157), 1 have the pleasure to 

 communicate to you that I have a collection of fourteen cocoa- 

 nut pearls (one of them I myself found in 1866 at Holontalo, 

 North Celebes, in the endosperm of the seed of the cocoa-nut) ; 

 two melati pearls {[asminium samhac) ; one tjampaka pearl 

 (Michelia longifolia), found in the flowers, according to the 

 natives. One of the cocca-nut pearls has a pear-shaped form, 

 the length being 28 mm. The common name amongst the 

 natives for this kind of pearls is mustika. 



Utrecht, September 6. J. G. F. Riedel, 



STARS WITH REMARKABLE SPECTRA. 

 I. 



No. 152 Schjellerup {D.M. + 46° No. 1817). 

 Place i887'o, R.A. \2h. 39W. 47^., Decl. 46° 3'-5 TV. 



nrmS star, No. 290 in Mr. Birmingham's Catalogue of 



*■ Red Stars, may very fittingly be taken as a sample 



of the stars possessing spectra of the fourth type, to use 



Secchi's nomenclature, or of the second division of the third 



type, to follow Vogel's — spectra, that is, in which the pro- 

 minent feature is a series of dark bands alternating with 

 bright spaces, and in which the dark bands are, as a rule, 

 sharp and dark on the less refrangible side, or that nearer 

 the red, but which gradually fade away into nothingness 

 on the more refrangible side, or that towards the violet. 

 The present star, though not perhaps the one in which 

 the series of bands is most completely developed, has yet 

 a spectrum which is a very beautiful example of the type; 

 the bright interspaces, or zones as they are technically 

 called, are vivid and striking, and the bands broad and 

 dark, and it possesses the additional advantage that, 

 though only of magnitude 5*5, it is yet the brightest star 

 of its class in the northern heavens. 



The purpose of the accompanying diagram, in which 

 the spectrum of 152 Schjellerup is seen side by side with 

 that of Tebbutt's comet of 1881, and with a particular 

 carbon spectrum, is to bring into prominence the meaning 

 of the remarkable series of shaded bands which charac- 

 terize it. In 1869, Secchi had declared that these bands 

 coincided as to position with the bands of the carbon 

 spectrum ; but, as Dr. Huggins shortly after stated that 

 he had compared the spectrum of carbon with that of a 



Scale of Wave-Lengths. 



red star, and found that the two did not coincide, it was 

 generally assumed that the Italian observer was mis- 

 taken, the well-known skill and accuracy of the great 

 English spectroscopist rendering it very unlikely that his 

 observation should be in error. As the event proved, 

 both were right ; it was only the natural inference that 

 the two observations were contradictory that was at fault. 

 Our knowledge of the beautiful and complicated spectra 

 of carbon had not then attained its present precision, and 

 it escaped remark that the spectrum with which Secchi 

 had compared the red stars was not the same that 

 Huggins had used for that purpose. Even now spectro- 

 scopists are not wholly unanimous as to whether we 

 should regard these two spectra as both belonging 

 to elemental carbon at different temperatures, or as 

 belonging to two different classes of carbon compounds — 

 those with oxygen and those with hydrogen. The 

 spectrum which Secchi had used was that which, accord- 

 ing to Thal^n and others, characterizes the hydrocarbons ; 

 whilst Huggins used that of the oxides. 



The former spectrum is one which was already of high 

 importance to the astronomer. Huggins had shown, in 

 1868, by comparing Winnecke's comet with olefiant gas, 



that the three bright bands so typical of a comet coin- 

 cided precisely with this form of the carbon spectrum ; 

 and now Duner and Vogel have placed it beyond a doubt 

 that in the spectrum of the red stars we see the same 

 spectrum, only reversed — an absorption instead of an 

 emission spectrum. The agreement as to the place of 

 the sharp, well-marked, less refrangible edge of the three 

 principal bands — the yellow, the green, and the blue — 

 is exact within the limit of errors of observation ; the 

 shading-off towards the violet is similar in character, and 

 there are indications of the presence of some at least of 

 the secondary flutings which in the carbon spectrum fol- 

 low the great leaders of the bands in so charmingly 

 rhythmical a manner. The orange band also, placed in 

 a fainter part of the spectrum, and so more difficult to 

 observe, is present, there can be little doubt, in the ab- 

 sorption spectrum of the red stars, though its bright 

 analogue has seldom been satisfactorily traced in the 

 spectrum of a comet ; the violet band, on the other hand, 

 appears to have been better seen in the comet than in the 

 red star. 



The following table will show the character of the corre- 

 spondence of the principal bands of the three spectra — 



