NO. O SAMUEL PIERPONT LANGLEV — ABBOT 5 



it will appear subsequently can hardy equal o".3 in diameter, and 

 are probably less. (Secchi is the only observer, as far as I know, who 

 appears to have seen and measured them. He observed them in the 

 edges of the pores, and reckons their size at ".J^ to "^, but does not 

 estimate their number or point out their relations to the ' rice-grains.') 

 They are irregularly distributed, with a tendency to aggregation in 

 little clusters (the clusters being the rice-grains), and their existence 

 accounts for the diversity and irregularity in the outline of the latter, 

 Mr. Huggins has acutely remarked upon, while it of course makes 

 clear the reason of the apparent increase in the number of ' rice- 

 grains ' with increasing telescopic power. 



" We are now prepared to study the minute structure of the photo- 

 sphere under another aspect, as it appears in the spots. It is impos- 

 sible to make such a drawing as that here given from any single de- 

 lineation, owing to the rapidity with which spots change their form. 

 I have accordingly, while taking the general contour and many details 

 from drawings of the great spot of March 5 and 6, 1873, added the 

 results of numerous studies of detail in other spots, made during the 

 past two years 



" To represent the gradations of light from the intensest splendor 

 to the darkness of the nuclei, we have here only the limited range 

 between a white and a black pigment. This almost compels partial 

 falsity in the degrees of shade, and there is, for instance, in the draw- 

 mg, a relative exaggeration of the shade which marks the outer 

 boundary of the penumbra, and without which the important details 

 would be hardly visible. 



" It is practically impossible, in the brief intervals of perfect defini- 

 tion during which such work can be carried on, to so multiply mi- 

 crometric measurements, that from their concordance any idea of 

 their probable error is obtainable by the usual treatment. Measure- 

 ments taken at different times, and on different parts of the penumbra, 

 by counting the number of filaments in a given space, give from 

 o".7 to i".o as the average distance from center to center of parallel 

 filaments separated by scarcely measurable intervals ; at the same time 

 that the distance in some parts is greater, it is in others much less. 



" Solar cyclones, which, even without the aid of the spectroscope, 

 we see are incomparably more violent than our own tropical tornados, 

 act on the filaments without destroying their identity. It is probable 



