RUMBX^RD SPECTROHELIOGRAPH, 141 



llocciiltis l)c slunvii on such a ph()t()<iTaph, since the line they produce 

 is not l)r()ii(I enongrh to enter the second slit. If the slit were set 

 nearer to the center ()f the line the photograph should repi-esent 

 a section of the flocculus corresponding to a higlier level, where a 

 narrower line is produced. It is evident that while none of the 

 higher and rarer calcium vapor could he shown in this photograph, 

 it might nevertheless include regions lying below it, where the 

 calcium vapor is dense enough to produce a broader line. However, 

 since the calcium vapor is rising from a region of high pressure to 

 one of a much lower pressure, it nuist expand as it rises, and there- 

 fore a section at any level should, in general, l)e of a larger area 

 than a section of the same flocculus at any lower level. As a con- 

 sequence of the increasing extent of the vapor with the altitude, and 

 the increase of brightness observed when passing from K^ to K^, 

 a photograph corresponding to a given level is not necessarily af- 

 fected in any considerable degree by the existence of the denser vapor 

 below, except in cases where the high-level va})or does not lie inune- 

 diately above the low-level vapor. Low-level phenomena, even 

 when very bright, may be wholly concealed l)y general excess of 

 radiation, or in some cases by absorption, of the calcium vapor at 

 high levels. Moreover, it is of course to be understood that the 

 term '' level " is not used here in a strict sense. A section of a 

 large flocculns jihotographed vvith K. might, for example, correspond 

 to a much greater height above the photosphere than that of the 

 minute fiocculi shown on the same photograph. It must never be for- 

 gotten, when examining the photographs, that composite effects 

 are very likely to be present. 



Such considerations regarding the possibility of photographing 

 sections of the fiocculi at diHerent levels are borne out by the photo- 

 graphs, as will be seen by reference to the accompanying illustra- 

 tions. Plates VII and viii represent the spot group of 1903, April 

 29, as photographed with four different settings t>f the second slit. 

 These were taken within such time limits and in such an order tliat, 

 as no distinctly eruptive phenomena were present, the ])rincipal dif- 

 ferences between tlie photographs are therefore to be attributed to 

 differences in the extent and brightness of the vapor at various levels, 

 and not to changes going on in the sun at the time. 



The assumption that these photographs represent sections of the 

 calcium fiocculi at different elevations seems to be the simplest and 

 most satisfactory way of explaining the results obtained. Essen- 

 tially conclusive evidence in favor of this assumption is afforded by 

 the photographs of the recent great sun spot. (Pis. xii to xviii.) 



With the aid of this additional means of research we may return 

 to a consideration of the structure of the fiocculi. It has already 

 been remarked that the general surface of the sun appears to be cov- 



