t)6 Observations of a Comet, 



this been the case, the deficiency, T think, would have been 

 perceived, notwithstanding the smallness or' the object. 

 Those who are acquainted with my experiments on smalls 

 silver, globules-* will easily admit, that the same telescope 

 which could show the spherical form of balls, that subtend- 

 ed only a few tenths of a second in diameter, would surely 

 not have represented a- cometary disk as circular, if it had! 

 been as deficient as are the figures which give the calculated 

 appearances. 



If these remarks are well founded, we are authorised to 

 conclude, that the body of the comet on its- surface is sclf- 

 luminous, from whatever cause this quality may be derived.. 

 The vivacity of the light of the comet also had a much 

 greater resemblance to the radiance of the stars, than to the 

 mild reflection of the spin's beams from, the moon, which i* 

 an additional support of our former inference. 



The changes- in the brightness of the small stars, when 

 they are successively immerged in the tail or coma of the 

 comet, or cleared from them, prove evidently, that they are 

 sufficiently dense to obstruct, the free passage of star-light. 

 Indeed if the tail or coma were composed of particles that 

 reflect the light of the sun, to make them visible we ought 

 rather to expect that the number of solid reflecting parti- 

 cles, required for this purpose, would entirely prevent our 

 seeing any stars through them. But the brightness of the 

 head, coma, and tail alone, wilbsufficiently account for the 

 observed changes, if we admit that they shine not by reflec- 

 tion, but by their own radiance; for a faint object projected 

 on a bright ground, or seen through itj will certainly appear 

 somewhat fainter^ although its rays shointf meet with no ob- 

 struction in coming to the eye. Now, as in this case we are 

 sure of ths bright interposition of the parts of the comet,, 

 but have no knowledge of floating particles, we ought cer- 

 tainly not to ascribe an effect to a hypothetical caus^, when 

 the existence of one, quite sufficient to explain the pheno- 

 mena, is evident. 



If we admit that the observed full illumination of the disk 



* Phi!osoph"-al Transactions for ISOj, p. 58,. the. 5th experiment. 



of 



