L. Boss—Tail of Comet b, 1881. 309 
He refers the tails of comets to three general types, distin- 
guished by the value of (1—y) employed in their theoretical rep- 
resentation. The value of (1—y) (expressed in the Newtonian 
unit) for Type I is 11:0 to 12:0; for Type II, about 1°3 ; for Type 
IIT, 0°3, or less. The value of (1—y) for Type II, however, is 
found to vary considerably in different cases without losing 
its distinctive character. It is possible to introduce the effect 
due to the initial velocity of projection from the nucleus,.and 
this, of course, modifies the value of (1—y) which would other- 
wise be assumed. This effect will evidently be proportionally 
least in tails of Type I, and will increase in importance as the 
projected from the nucleus equally in all directions with equal 
velocities, the effect will be mainly shown in the breadth of the 
tail. Thus we invariably find tails of Type I to be narrow in 
comparison with those of Type IJ,—a fact which finds satisfac- 
tory explanation in the relatively small effect, which would be 
produced by the action of initial velocity, when the repelling 
force is relatively very great. But since cometary emissions 
appear to take place mostly on the side of the nucleus nearest ° 
the sun, the assumption of the value zero for initial velocity 
will always render the value of (l—y) computed from observa- 
tion, too small. 
It will be interesting to examine our observations of the tail 
of comet b 1881, with a view to determining to what extent 
they conform to the normal types. In a preliminary discussion 
like this, which is founded on few observations of small weight, 
it will not be worth while to include the effect of initial velocity 
of emission, When a great number of observations of the tail 
heim, previously cited. 
Let : 
M = Date when a given particle is observed in the tail. 
= Time of emission of that particle from the nucleus. 
M"= Perihelion passage of the particle. 
E = Eccentricity of the hyperbolic orbit. 
I= Angle between the radii vectores of the partiele and nucleus 
at the time M. For the particle referred to the nucleus, 
this angle will evidently always be retrograde to the 
motion of the nucleus. > 
4 = Distance of the particle from the nucleus at the time, M. 
