138 RtTMFOKD SPEOTROHELIOGRAPH. 



siderable oi)enings caused by the absence of the white nodnles at certain point?, 

 and the consequent exposure of the gray medium which forms the general buck- 

 gi'ound. These ojienings have been called pores; their variety of size makes 

 any measurements nearly valueless, though we may estimate in a very rough 

 way the diameter of the more conspicuous at from 2" to 4". The bright 

 nodules are themselves not uniformly bright (some being notably more brilliant 

 than their fellows and even unequally bright in portions of the same nodule), 

 neither are they uniform in shape. They have just been spoken of as i-elatively 

 definite in outline, but this outline is commonly found to be irregular on minute 

 study, while it yet affects, as a whole, an elongated or oval contour. Mr. Stone 

 has called them " rice grains," a term only descriptive of their appearance with 

 an aperture of 3 to 4 inches, but which I will use provisionally. It depicts their 

 whiteness, their relative individuality, and their approximate form, but not their 

 irregular outline, nor a certain tendency to foliate structure which is character- 

 istic of them, and which has not been sufficiently remarked upon. This irregu- 

 larity and diversity of outline have been already observed by Mr. Huggini-'. 

 Estimates of the main size of these bodies vary very widely. Probably Mr. 

 Huggins has taken a judicious mean in averaging their longer diameter at 1".5 

 and their shorter at 1", while remarking that they are occasionally between 2" 

 and 8", and sometimes less than 1", in length. 



* * * In moments of rarest delinitiou 1 have resolved these "rice grains" 

 into minuter components, sensibly round, which are seen singly as points of light, 

 and whose aggregation produces the " i-ice-grain " structure. These minutest 

 bodies, which I will call granules, it will appear subsequently, can hardly equal 

 0."3 in diameter, and are probably less. 



* * * It seems to me that there is no room for doubt that " filaments " and 

 " granules " are names for different aspects of the same thing : that filaments in 

 reality are floating vertically all over the sun, their upper extremities appearing 

 at the surface as granules, and that in spots we only see the general structure 

 of the i)hotosphere. as if in section, owing to the filaments being here inclined. 



* * * Speaking without reference to spectroscopic investigations. It seems 

 to me that we h.-ivc in the behavior of our filaments a presumption as to the 

 existence of ascending currents in the outer peinnnbra, and of both ascending 

 and descending currents at the umbral edge; ascending ones being the more 

 usual. 



An examination of the minute calcium flocciili photographed vith 

 the Tvumford spectroheliog-raph will show that they closely reser.iltle 

 the photospheric " grains " described by Langley and illustrated in 

 Janssen's photographs. Plate vi is reproduced from one of our nega- 

 tives on the scale cliosen for the majority of the photographs in Vol- 

 ume T of the ]\reud<m Annals. This i)hotograph was made with tlie 

 slit set at the center of the H line on a day when the seeing was par- 

 ticularly good. In fig. 3, pi. iii, squares 10 seconds of arc on a 

 side are shown. These permit of an accurate determination of the 

 size of the individual elements of the structure. Measurements made 

 on our best negatives show that the minute calcium flocculi range in 

 diameter from less than 1 second to several seconds of arc, thus cor- 

 responding closely with the " grains " of the photosphere. 



On ti>e working liy])othesis at present employed to interprel llie 

 results obtained Avith the llumford spectroheliograph, it is considered 



