Nov. 27, 1879] 



NATURE 



83 



the temperature be raised equally and the volume retained 

 at its original value, no deposition takes place. Those 

 experiments have been done with such solvents as alcohol 

 (ethyl and methyl), ether, carbon disulphidc and tetra- 

 chloride, paraffins, and defines, and such solids as 

 sulphur, chlorides, bromides, and iodides of the metals, 

 and organic substances such as chlorophyll and the 

 aniline dyes. Some solutions show curious reactions at 

 the critical point. Thus ethyl alcohol, or ether, deposits 

 ferric chloride from solution just below the critical point, 

 but re-dissolves it in the gas, when it has been raised 8° 

 or 10° above that temperature. 



It appeared to us to be of some importance to examine 

 the spectroscopic appearances of solutions of solids when 

 their liquid menstrua were passing to the gaseous state, 

 but as all the substances we have yet been able to obtain 

 in the two states give banded spectra with nebulous 

 edges, we are only able to state that the substance does 

 not show any appreciable change at the critical point of 

 its solvent. Such was the case with anhydrous chloride 

 of cobalt in absolute alcohol. It was suggested to us by 

 Prof. Stokes that the substance obtained by the decom- 

 position of the green colouring matter of leaves by acids, 

 and which yields a very fine absorption spectrum, might 

 be useful for our purpose. We have prepared the sub- 

 stance according to the careful directions so kindly fur- 

 nished us by Prof. Stokes, and find that it shows the 

 phenomenon in a marked manner, whether dissolved in 

 alcohol or ether. The compound is easily decomposed 

 by heat under ordinary circumstances, and yet can be 

 dissolved in gaseous menstrua, and raised to a tempera- 

 ture of 350 without suffering any decomposition, showing 

 the same absorption spectrum at that elevated tempera- 

 ture as at 1 5 . 



We considered that it would be most interesting to 

 examine by this method a body such as sodium, which, 

 besides being an element, yields in the gaseous state 

 sharp absorption lines. An opportunity seemed to be 

 afforded by the blue solution of sodium in liquefied 

 ammoDia, described by Gore, 1 but we found that, on 

 raising the ammonia above its critical point, the sodium 

 combined with some constituent of the gas, forming a 

 white solid, and yielding a permanent gas, probably 

 hydrogen. 



There seems, in some cases, to be a slight shifting of 

 the absorption bands towards the red, as the temperature 

 rises, but we have as yet been able to make no accurate 

 measurements. 



When the solid is precipitated by suddenly reducing 

 the pressure, it is crystalline, and may be brought down 

 as a '' snow " in the gas, or on the glass as a " frost," 

 but it is always easily redissolved by the gas on increasing 

 the pressure. These phenomena are seen to the best 

 advantage by a solution of potassic iodide in absolute 

 alcohol. 



We have, then, the phenomenon of a solid with no 

 measurable gaseous pressure, dissolving in a gas, and not 

 being affected by the passage of its menstruum through 

 the critical point to the liquid state, showing it to be a 

 true case of gaseous solution of a solid. 



Private Laboratory, Sword Street, Glasgow 



OX PHOTOGRAPHING THE SPECTRA OF THE 



STARS AND PLANETS - 

 T7 OR many years it has seemed probable that great 

 -*- interest would be attached to photographs of the 

 spectra of the heavenly bodies, because they offer to us 

 conditions of temperature and pressure that cannot be 

 attained by any means known at present on the earth. 

 The especial point of interest is connected with considera- 



1 Prcc. Roy. Soc, vol. xxi. p. 145. 



2 Read before the National Academy of Sciences, October 2S, by Henry 

 Draper, M O. ' 



tions regarding the probable non-elementary nature of the 

 so-called elementary bodies. There has long been a sus- 

 picion in the minds of scientific men that one or more 

 truly elementary bodies would be found from which those 

 substances which have not as yet been decomposed are 

 formed. The recent publications of Lockyer have at- 

 tracted particular attention to this topic. 



The most promising laboratory processes for accom- 

 plishing the dissociation of our present elements depend 

 upon the action of heat, especially when accompanied by 

 electrical influences, and upon relief of pressure. But 

 the temperature we can employ is far below that found in 

 the stars, which is comparable only with the heat of our 

 sun, and when in addition the application of heat is 

 restricted by the narrow range of circumstances under 

 which we can also reduce the pressure, complete success 

 seems to be impracticable in the laboratory. 



But in the stars, nebuke, and comets, there is a multi- 

 tude of experiments all ready performed for us with .1 

 variety of conditions of just the kind we need. It re- 

 mains for us to observe and interpret these results, and 

 this is the direction I have sought to pursue. 



There is but one mode of investigation that can add 

 materially to the knowledge astronomy has given us of 

 the heavenly bodies — that is the spectroscopic. This in 

 its turn is capable of a subdivision into two methods, one 

 by the eye, the other by photography. Each of these 

 has its special advantages and each its defects. The eye 

 sees most easily the middle regions of the spectrum, and 

 can appreciate exceedingly faint spectra ; by the aid of 

 micrometers it can map with precision the position of the 

 Fraunhofer lines, and by estimation it can with tolerable 

 accuracy approximate to the relative strength, breadth, 

 and character of these lines. The character of the spec- 

 trum lines is, however, of great value for the purposes 

 we are now speaking of, and the greatest precision is 

 needed. Photography, on the other hand, as applied to 

 faint spectra, deals mainly with the more refrangible 

 region, and cannot at present be employed in stellar work 

 below the line F. Fortunately there is no break in the 

 spectrum between the place where the eye leaves off and 

 photography begins, and hence the two methods lend one 

 another mutual assistance. The photograph, when suit- 

 ably accommodated with a standard reference spectrum 

 from some known source, gives valuable indications as to 

 the positions and all the peculiarities of the lines. 



But the application of photography to the taking ot 

 stellar spectra is surrounded by obstacles. These are 

 partly due to the small quantity of light to be dealt with, 

 and partly to the fact that it is necessary to overcome the 

 motion of the earth and other causes, such as atmospheric 

 refraction, which seem to make a star change its place 

 continually. The exposures of the sensitive plate require 

 to be sometimes for two hours, even with a large tele- 

 scope; and if during that time the image of the star at 

 the focus of the telescope has changed place 3 Jrj of an 

 inch, the light no longer falls on the slit of the spectro- 

 scope. The changes of the earth's atmosphere in regard 

 to photographic transparency, as well as by fog, also offer 

 impediments and promote the chances of failure. There 

 is often a yellow condition of the air, which may increase 

 the length of exposure required forty times or more. 



It will from what has been said above, be readily per- 

 ceived that a research such as this consumes a great deal 

 of time ; in fact, these experiments and the preparations 

 for them have extended over more than twelve years. A 

 large telescope is required, and for many reasons the 

 reflector at first seems most suitable. Recently, however, 

 I have found that the refractor has also some special 

 advantages. 



In 1866 I had already constructed a silvered glass 

 reflector of 15J inches aperture, which was commenced 

 in 1858, and had taken with it many hundreds of photo- 

 graphs of the moon. But as the mounting had been 



