W. A. Norton on Comets. 95 
For the first term gives the parabolic form, and the subtractive 
term varies as aes which is also the law of variation of the dis- 
tance between two parabolas having a common focus and axis, 
estimated in a direction perpendicular to the tangent. The onl 
essential difference between this case and that in which the repul- 
sion is in excess, is that the returning particles which pass in the 
vicinity of the nucleus will now be attracted by it, and pursue 
more concave toward the nucleus, instead of convex. 
A detailed discussion can therefore alone settle definitively the 
question between these two possible modes of development of the 
of a comet from its nucleus. This must be left for another 
It consists, a parently, of two jets of luminous matter, issuin; 
from the xe and convex toward each other, like the two 
branches of a hyperbola. Upon the theory of no ge it may 
be explained by the intersection of the orbits © — at 
pass in the immediate vicinity of the nucleus. ose which 
make the closest approach to the nucleus, wi 4. 
most deflected from their course. The consequence 
Will be that the trajectory of each particle will 
intersect those of all the particles that come be- 
tween it and the nucleus. This curious result is 
ustrated in fig. 4. It is a case similar to that of ~ 
Caustic curves in optics. 
It will be seen on referring to fig. 8, that the par- 
occasion to observe that as the angle of emis- 
Sion becomes less, the orbits, as they traverse the head, and are 
about leaving it, lie nearer together; and rathersuddenly 
Tate from each other, when the angle becomes small. This 
ms me tegen A indicated in fig. 3. It may be inferred from t 
fact that the parameter of the parabolic path of the cle con- 
“inues to increase as the angle of projection diminishes, until it 
quite small. The liarity here noticed affords an 
¢xplanation of the sudden curved terminations of the luminous 
; disc that the central portion of the head of Donati’s 
, Omet, as seen in te in Hig Pope Sarl ped 
appearances (e. g. Halley’s Comet as seen truve, 
Oct. 29th, 1885, and the bite on et 92). The natural result, 
> 
