Deceptive Figures. 25 



sphere might be assumed to be from 450 to 500 miles.* Oar 

 positive evidence is not, perhaps, very strong. The circum- 

 stance that the constitution of an atmosphere so far removed 

 admits of determination by spectrum-analysis seems to indicate 

 that the height of the atmosphere must be considerable. Mr. 

 Grover's observation of a penumbra surrounding the planet's 

 shadow on the ring [Astronomical B.egister for August, 1865, 

 p. 212), "always noticed" even with the moderate telescopic 

 power applied by this observer, also points to the existence of 

 a strongly refractive atmosphere. Herschel's observation of 

 a remarkable retardation of the occultations of Mimas and 

 Enceladus has not been confirmed; but, perhaps, this is 

 hardly to be wondered at. There are not ten telescopes in 

 the world capable of following these two satellites up to the 

 moment of their disappearance, and since the date of Herschel's 

 observation there have only been five intervals, each of a few 

 weeks only, during which the observation has been possible. 

 It is true that where the further part of the ring appears to 

 meet the disc at an acute angle, its outline should be distorted 

 if the planet have a refractive atmosphere ; but it is easy to 

 see that the distortion would scarcely be appreciable even with 

 the most powerful instruments, and Mr. Grover's observation 

 above recorded is sufficient to show, that details observable 

 with very moderate powers may for a long time escape obser- 

 vation. It will presently be seen that I am not here losing 

 sight of the evidence pointing to a reflective power possessed 

 at times and in certain latitudes by Saturn's atmosphere. 



Now it has been observed by Professor Challis,t that the 

 atmospheres of planets must have certain definite limits, since 

 "the density«continually decreasing with the height, a point 

 must at length be reached at which the upward repulsive force 

 of an atmospheric stratum is just equal to the force of gravity ; 

 in which case there can be no downward repulsive force, and 

 therefore no further extension of the atmosphere." And he 

 considers, that the effect of an atmosphere on our estimate of a 

 planet's diameter will depend on "the relation between the 

 gradation of density of the atmospheric strata and the curva- 

 ture of the globe." This relation may be such that a ray 

 would "pass through the atmosphere in a course which grazes, 

 or is a tangent to, the interior globe ;" or else, that rays could not 

 reach " the surface of the globe in a direction making with it 

 an angle less than " a certain limiting angle. In the first case 



* Sehroter estimated the height of the atmosphere of Ceres at 668 miles (!) ; 

 a result which can hardly be considered as established. 



t In an article on the " Indications, by Phenomena, of Atmospheres to the 

 Sun, Moon, and Planets," Notices of the Astronomical Society, vol. xxiii., pp. 

 230-238. 



