350]
ON THE CLASSIFICATION OF CUBIC CURVES.
363
the inflexion, a line which cuts harmonically the chords through the inflexion, and which
when the inflexion is at infinity becomes a diameter. The remarks previously made as
to osculating asymptotes apply therefore to diameters, viz. the Hyperbolas may have no
diameter, a single diameter, or three diameters, &c.
40. Newton speaks also of the “ centre ” of a cubic curve; viz. there may be a
point on the curve such that for any line through this point the two radius vectors
are equal and opposite to each other, or that the sum r + r' of the two radius vectors
is = 0. The centre is in fact a point of inflexion which has for its polar the line
infinity. The curves which may have a centre are the Hyperbolas (redundant or defective),
the Central Hyperbolisms (of the hyperbola or ellipse) and the Cubical Parabola. For
the hyperbolas, the three asymptotes and the satellite line must meet in a point of
the curve, which point is then the centre; for the central hyperbolisms the onefold
asymptote and the satellite line must meet in a point of the curve, which point is
then a centre; and for the cubical parabola no condition is required, but the inflexion
is a centre. I remark here, in passing, that the notion of a centre as just explained
has no place in Plucker’s Classification, and that the two Newtonian species 58 and 59
(hyperbolisms of the hyperbola) and the two Newtonian species 61 and 62 (hyperbolisms
of the ellipse) which differ, the two of a pair from each other, according as there is
no centre or a single centre, form each pair a single species with Plticker; viz. they
are 198 and 201 respectively.
41. It has been already remarked that the three asymptotes of a Hyperbola may
meet in a point. As to this it is to be noticed that from any point we may draw
six tangents to a cubic, the points of contact lie on a conic, the conic polar of the
point: if, however the point lie on the Hessian of the cubic, then the conic breaks
up into a pair of lines, each of which is a tangent to the Pippian; the two lines
meet in a point of the Hessian, which point forms with the first mentioned point a
pair of conjugate poles of the cubic( J ).
Conversely, any tangent of the Pippian meets the cubic in three points, the
tangents at which meet in a point of the Hessian; and from this point we may draw
to the cubic three other tangents the points of contact of which lie on a line which
is also a tangent of the Pippian, and the two tangents of the Pippian meet in a
point of the Hessian; the two points of the Hessian being conjugate poles of the
cubic. In particular, if the line infinity is a tangent of the Pippian, then the three
asymptotes meet in a point of the Hessian, and the three tangents from this point to
the cubic touch the cubic in three points lying on a line which is a tangent of the
Pippian, and which meets the line infinity in a point forming with the first mentioned
point a pair of conjugate poles of the cubic.
I proceed now to explain the classification of Newton so far as relates to the
division into genera, and the classification of Pliicker so far as relates to the divisions
immediately superior to the groups.
1 See as to this theory my Memoir on Curves of the Third Order. Phil. Trans, p. 147 (1856), [145].
46—2