3(» 



KNOWLEDGE. 



[February, 1903. 



which, translated into English, becomes : — 



" Above the highest summit art figured a cross, guardian 

 of the city." 



Messrs. Lethaby and Swainson remark that " i'ypa^t 

 leaves no doubt that a mosaic cross on the interior is 

 intended."* It is quite clear, however, that the expression 

 " ahooe the highest summit " could never apply to an 

 inner cross in mosaic. A real picture is always on the 

 surface of the material on which it is depicted ; it can 

 neither be above nor below that surface. Hence no mosaic 

 cross, visible from the floor, could ever have been above 

 the spherical surface on which it was drawn ; it must 

 necessarily have been on that surface. But in the case of 

 a solid cross, erected on the outside, the above is welcome, 

 inasmuch as the cross then rises over the outer surface of 

 the dome. These arguments render impossible the 

 meaning of painted to typx<fie, as, in that case, the painting 

 would have to be effected on empty air, a few feet above 

 the top of the cupola. An English equivalent to /yp«^£, 

 applicable to a solid cross cast in metal, is figured ; and, 

 with its adoption, everything is plain sailing. Another 

 argument, but of an indirect bearing on the point at issue, 

 is that a cross on the dome of a great cathedral would 

 guard the city somewhat indifferently if painted in the 

 darkness of a vault. The watch would be much more 

 effective if the guardian could be given a chance of seeing 

 what he was guarding, and be seen from it ; the people 

 would rejoice in the idea of seeing their palladium from a 

 distance in the city. 



A careful synthesis of historical facts will enable us to 

 undertake a fairly trustworthy restoration of the outer 

 appearance of St. Sophia at one of the most flourishing 

 periods of Byzantine history, at the time of Basil the 

 Macedonian, in the ninth century. Adding the bell tower 

 erected here in 865 by the emperor Michael in order to 

 receive the bells presented to him by the Doge of Venice, 

 and actually seen in position by Grelot towards 16"9 ; 

 restoring the steps of the atrium ; liberating the western 

 arch and semi-dome of their additional thicknesses ; also 

 giving the west wallf and the dome their original appear- 

 ance, we Tmmistakably launch on something like Fig. 2, 

 where the only weak point lies in our ignorance as to tiie 

 number of arched openings piercing the atrium wall on 

 the west side. A similar treatment of the south facade, 

 but for a more recent date, led the writer to the elevation 

 shown on Fig. 3. Apart from the omission of the 

 buttress masses erected in 1317, this elevation differs 

 from the present view of the mosque in some other par- 

 ticulars, such as the saddle roofs of the little chambers at 

 the buttress tops ; the restoration of a room at the top of 

 the staircase on the south-east angle of the great platform ; 

 and the insertion of two buttress arches reaching up to 

 the windows of the cupola. It was during the 1847- 

 1849 restoration that the roofs of the chambers were 

 rendered cylindrical ; that the corner room and buttress 

 ai-ches were removed, the office of the latter being taken 

 advantageously by a double cincture of iron round the 

 base of the dome. 



* Ibid, p. 42, note. 



t It was thought safe to butti-ess the dome on the west at an 

 unknown period ot the church's history, and this was done by length- 

 ening radially thirteen of the piers between the upper windows. 

 But there was not enough space between the edge of the central 

 western piers and the edge of the west wall, so that the work entailed 

 a thickening of the latter. This thickening is, therefore, not original, 

 and it does not exist to the east, so that its representation in "original 

 forms " of the church by M. Choisy, as well as by Messrs. Lethaby 

 and Swainson, is doubtless attributable to some sort of error. 



THE CHEMISTRY OF THE STARS. 



I. — INTRODUCTORY. 



By A. FoWLEK, F.R.A.S. 



It is a striking fact that more is known with certainty as 

 to the substances which help to build up the sun, and the 

 still more distant stars, than of the materials which exist 

 at depths greater than a few miles below the surface of 

 our own planet. As to the interior of the earth, we can 

 in no way examine any portion of it, and can only speculate 

 as to what may be there by reasoning from such facts as 

 the known increase of density as the centre is approached. 

 In the case of the heavenly bodies, however, we have at least 

 the advantage of being able to see them, and it is in fact 

 upon the spectroscopic analysis of the light which they 

 radiate into space that our knowledge of their chemical 

 composition is founded. 



The general principles involved in the identification of 

 the substances composing the heavenly bodies may be 

 summed up in a few words. An incandescent gas or 

 vapour sends out rays of light having definite wave- 

 lengths, whether it be raised to incandescence in a labora- 

 tory experiment or by the conditions to which it is sub- 

 jected in sun, star, or nebula ; so that all that seems 

 necessary is to find by trial which of the substances open 

 to experiment emits rays corresponding with those of the 

 body under investigation. This analysis of light is of 

 course carried on by the use of a prism or diffraction 

 grating, by which means the rays are dispersed into their 

 component elementary waves, and as the light is usually 

 first passed through a narrow slit, each of these com- 

 ponents is represented by a " spectrum line '' of definite 

 colour and position. An interesting example of this 

 process is illustrated by Fig. I, which shows how the 

 presence of helium in Nova Persei, at one stage of its 

 history, was established by comparing its spectrum with 

 that of helium enclosed in an ordinary vacuum tube, and 

 rendered luminous by electric discharges. ,_, 



{To be continued.) 



Fia. I.— Spectrum of Nora Persei, lltiU, August 11, with Helium 

 comparison, showing coincidence of the D3 line. (Lick Observatory.) 



IJsually, however, the direct radiations of the substances 

 composing the heavenly bodies cannot be observed. Thus, 

 the light of the sun, and of the great majority of the 

 stars, exhibits dark lines on a coloured background when 

 analysed by the prism, but the far-reaching discovery by 

 Kirchhoff that incandescent gases and vapours absorb 

 light of the same wave-lengths as they themselves emit at 

 the same temperature, shows that the dark lines are just 

 as distinctive as the corresponding bright ones, so that 

 the chemical origins of dark lines may be investigated by 

 the process of matching employed in the case of bright 

 lines.* An example of this is given in Fig. 2, which 

 shows how the presence of iron in the sun is demonstrated 

 by the coincidence of numerous dark solar lines with as 

 many bright ones of the metal, the source of light in 

 the latt<ir case being an electric arc lamp having poles 

 consisting of iron rods. 



* A general summary of the principles of spectrum analysis, and 

 an account of experiments illustrating them, will be found in Kxow- 

 LEDGE DiABT for the current year. 



