SOLAR ECLIPSE, MAY 6, 1883. 125 



appear as though arranged with reference to a disk, not a globe; thus the feature so puzzling in 

 the generally accepted theory is a necessary consequence of the views here urged. 



I have attempted to show in a paper on the physical constitution of the sun * that the pho- 

 tosphere is the locus of precipitation of some substance, very likely carbon, of very high vaporizing 

 l)oint; above this everything precipitated is, according to Kirciioff's principle, relatively dark. 

 It wilLprobablybe regarded as proved there, that the substance producing the general absorption 

 at the sun's limb is either liquid or solid matter in a state of fine subdivision. Granting this, it 

 follows that a particle immediately above the photosphere and within an ascending current, that 

 is, over a granule, will be illuminated chiefly from below. Moreover, in accordance with the argu- 

 ment developed in the paper cited and which is too long to insert here, the illumination would 

 be by light in which the Fraunhofer lines are necessarily very faint. Now, if the supposed parti- 

 cle is small enough, it will diffuse light which would be perfectly polarized in a direction at right 

 angles to the direction of illumination, and the plane of polarization would be that of reflection. 

 Perha))s the nearest possible experimental approach to the condition of an ascending column of 

 gas in the sun which the theory cited supposes, is the flame of a lamp when smoking. Such flauies 

 I have investigated, and find that the upper portion emits light very strongly polarized in the 

 direction anticipated. If, then, the constitution of the photosphere is as assumed, it follows that 

 just above it is a region where light is scattered tangeutially to the sun with strong radial polariza- 

 tion. According to the difl'raction theory the corona from such a sun must be strongly polarized at 

 its base, where this light is relatively strong, and continuously decrease outwards ; a conclusion 

 exactly contrary to the dicta of the old theory and in perfect accordance with observation. 



Thus, without contradicting any principle of physics, the difl'raction theory accounts for all 

 the observed features of the corona with a single exception, namely, its more minute structure. 

 Neither hypothesis succeeds here, at least if we accept all the curved forms indicated in the 

 engraved interpretation of the photographs. I will only add in this connection that every irreg- 

 ularity in the limb of the moon must cause an irregulai'ity in the light of a diffraction corona, 

 which, not improbably, may be of a wisp-like form; but at present both theory and experience 

 fail to give precise indications. 



j^ince the telegram announcing the results of the Eclipse Expedition was published I have 

 seen some statements that the visibility of the corona before second and after third contacts, and 

 the detection of the outline of Venus before it had entered on the sun or after it had i)assed ofl', 

 were definite proofs of the existence of an objective corona. I need hardly say that, in the case 

 of the moon, the fundamental principles of the theory oflered demands such a phenomenon. From 

 the visibility of Venus near the sun we can derive a very strong argument against the existence 

 of an objective corona, for, since the corona is strongest near the sun, that side of Venus nearest 

 the sun ought to be the most readily perceived. Instead of this being true, that side has never 

 been observed, though the more distant limb has frequently been seen.t 

 Yours, A'ery respectfully, 



CHARLES S. HASTINGS. 



* Proc. Amer. Acad. Sci., Amer. .Jour. Sci., Vol. , p. . 



tThe optical principles involved iu this phenoiueuou have been discussed iu the Sidereal Messeuj;er by the writer. 



