482 



NATURE 



[April iS, 1875 



by previous observers ; but of metalloids in this region I 

 have traced none. The persistency with which metal 

 after metal revealed itself to the exclusion of the metal- 

 loids led me to throw out the idea some time ago, that 

 perhaps the metalloids lay as a whole above the metals, 

 and shortly afterwards I obtained evidence which seemed 

 to me of a very satisfactory nature as to the existence of 

 carbon, its presence in the sun's atmosphere being ren- 

 dered probable by fluted bands, and not by lines. There 

 Avere two points, however, which remained to be settled 

 before the matter could be considered to be placed be- 

 yond all doubt. 



The first was to establish that the fluted bands gene- 

 rally present in the spectrum of the electric arc, as pho- 

 tographed, which bands vary very considerably in 

 strength according to the volatility of the metal under 

 experiment, were really bands of carbon — a point denied 

 by Angstrom and Thalen. 



This point I have settled by two photographs, in which 

 the carbon bands remain the same, though one spectrum 

 is that of carbon in air, the other of carbon in dry 

 chlorine. 



The next point Avas to insure accuracy by the most 

 positive evidence that there was absolutely no shift in the 

 carbon bands. Such a shift is produced when the part 

 of the arc photographed is not perfectly in the prolonga- 

 tion of the axis of the collimator of the spectroscope. Its 

 effect is to throw the lines of iron, for instance, a little to 

 the right or a little to the left of the Fraunhofer lines with 

 which they really correspond. 



I have now obtained a photograph which supplies such 

 evidence. There are metallic lines close to the carbon 

 bands Avhich are prolongations of Fraunhofer' s lines, 

 while the lines which I have already mapped at W. L.) 

 39'27 and 39*295, in the spectrum of iron, are also abso- 

 lute prolongations. Therefore there is no shift in the 

 carbon flutings, and the individual members of the fluted 

 spectra in the brightest portion are absolute prolongations 

 of a fine series of Fraunhofer lines in the ultra-violet. 



Now how does this connect itself with observations of 

 the upper parts of the solar atmosphere ? 



Angstrom has already shown that the true carbon h'nes 

 which we get Mhen a coil and jar are employed are 

 not reversed in the spectrum of the sun, and I have 

 already shown that the calcium spectrum in the sun 

 is similar to the spectrum obtained when the spark, 

 and not the arc, is employed. Accompanying the 

 change from a high to a higher temperature, there is 

 a change in the intensity of the lines — some thicken, 

 others become thinner. We can only match the relative 

 thickness of the solar calcium lines by employing a very 

 powerful coil and jar — so powerful, indeed, that the lines, 

 and not the flutings, of carbon would be visible in the 

 spark given by it. It is fair then to say that if carbon were 

 present with the calcium I'fi the sun's reversing layer, we 

 should get the lines of carbon when we get the calcium 

 lines appearing as they do. 



As we do not get this evidence, we are driven to the con- 

 clusion that the carbon vapourexists not onlyin a more com- 

 plicated molecular condition (as is evinced by the flutings) 

 than the metallic vapours in the sun's atmosphere, but at 

 a lower temperature. It must, therefore, exist above the 

 chromosphere, that is, in a region of lower temperature. 



Lower pressure, again, is indicated by the feeble reversal, 

 so that everything points to a high level. 



The question is, will this region be recognised during 

 the coming eclipse ? 



Coming down lower we reach a level better known, and 

 of which, perhaps, the interest during the eclipse will now 

 be less, if we except the possibilities opened out to us by 

 photography. One good photograph of the lines visible 

 in the lower chromosphere will be of incalculable value. 

 Attempts may be made on the cusps just before and 

 after totality, and if only one of these succeeds we 

 shall have the ordinary solar spectrum as a scale. If 

 good pictures near H can be secured, enough in- 

 formation now exists for that region to enable us 

 to determine the chemical origin of the bright lines 

 photographed. These remarks apply to attempts made 

 with spectroscopes furnished with slits in the ordinary 

 way ; there is little doubt, however, that the method 

 utilised for the Siam eclipse in 1875, the method 

 suggested by Prof. Young and myself for the Indian 

 Eclipse of 1871, will also be taken advantage of; here 

 the chromosphere itself becomes the slit. A dispersed 

 series of spectral images of the thing itself, instead of 

 the spectrum of a part of the image of it focussed on a slit^ 

 is obtained, the position of each image in the spectrum 

 enabling its chemical origin to be ascertained if only a 

 comparison spectrum can be secured at the same time. 



In 1875, in the expedition to Siam, the photographs 

 of this nature were obtained by means of a prism, and 

 the results obtained by that expedition led me to think 

 that, possibly, this method of using the coronal atmo- 

 sphere as a circular slit might be applied under very 

 favourable conditions if the prism, or train of prisms, 

 hitherto employed, were replaced by a reflection grating, 

 with which the generosity of Mr. Rutherford has made 

 many of us familiar, for the simple reason that while a 

 prism only gives us one spectrum, a brilliant grating 

 placed at right angles to an incident beam gives us 

 spectra of different orders, so-called, on each side of the 

 line, perpendicular to its surface. Of these two or three 

 are bright enough to be utilised on each side, so that we 

 can get six in all. 



To test this notion I made the following experiment 

 with a grating given to me by Mr. Rutherfurd. This 

 magnificent instrument contains 17,280 lines to the inch, 

 ruled on glass and silvered ; its brilliancy is remarkable. 



In front of the condenser of an electric lamp adjusted 

 to throw a parallel beam, I placed a circular aperture, 

 cut in cardboard, forming a ring some 2 inches in interior 

 diameter, the breadth of the ring being about \ inch. 

 This was intended to represent the chromosphere, and 

 formed my artificial eclipse. 



At some distance from the lamp I mounted a 3| inch 

 Cooke telescope. Some distance short of the focus 

 I placed the grating; the spectrum of the circular slit, 

 illuminated by sodium vapour and carbon vapour was 

 photographed for the first, second, and third orders on 

 one side. The third order spectrum, showing the exqui- 

 site rings due to the carbon vapour flutings, was produced 

 jn forty-two seconds. The first order spectrum, obtained 

 in the same period of time, was very much over-exposed. 

 It is, therefore, I think, not expecting too much that we 



