Elements of Solid Geometry-1891

Irving Strin^han
*th.
-4^ .C
NOTES
ON
ELEMENTS
OF
(ANALYTICAL)
SOLID GEOMETRY
BY
CHAS.
S.
VENABLE,
LL.D.,
PROFESSOR OF MATHEMATICS, UNIVE-RSITY OF VIRGINIA.
NEW YORK:
UNIVERSITY PUBLISHING COMPANY.
1891.
Math. 0ot.
IN preparing these Notes I have used the treatises of Gregory, Hymers, Salmon, Frost and Wostenholme, Bourdon, Sonnet et
Frontera, Joachimsthal-Hesse, and Fort
und Schlomilch.
C. S. V.
V
COPYRIGHT, BY
UNIVERSITY PUBLISHING COMPANY

1879.
NOTES ON SOLID GEOMETRY.

CHAPTER
I.
1. WE have seen how the position of a point in a plane with ref erence to a given origin O is determined by means of its distances In space, as there are three from two axes (Xr, Oy meeting in O.
So that each pair of axes dimensions, we must add a third axis Oz. determines a plane, CXr and Oy determining the plane xOy ; O.v
and O2 the plane xOz

tion of the point
its
Oy and Oz the plane yOz. And the posi with reference to the origin O is determined by
;
distances
PM, PN, PR from

measured on
This
the zOy, zOx,
xOy respectively,
these
distances being
Oz
respectively.
lines parallel to the axes CXv, Oy system of coordinates in space is called
and
The
System of Triplanar Coordinates, and the transition to it from the We can best System of Rectilinear Plane Coordinates is very easy.
conceive of these three coordinates of
corner of a parallelopipedon of which
P by
conceiving
as the
OA, OB,
OC
is
are the edges,
and the point P

2.
is
the opposite corner, so that
OP
one diagonal of
the equations
the parallelopipedon.
If
PM = OA =
P
are
a,
these c, and the point given by Measure on be found by the following construction equations may OX the distance OA = a, and through A draw the plane PNAR Measure on Oy the distance OB = 3, parallel to the plane yOz. and draw the plane PMBR parallel to xOz, and finally lay off OC The intersection c and dnuv the plane PMCN parallel to xOy.
of the point
PN = OB = a, y = b, z =
b,
PR = OC =
c,
of these three planes is the point P required. (Fig. i.)* Oz are called the axes of 3. The three axes Or, Oy,
respectively
*. ;
the
three planes xQy,

I.
x y, and z and yOz are called the xOz,

}
* For Figures see Plates
and
II.
at
end of book.
814023
planes xy, xz and yz respectively. The point whose equations are c is called the point (a, b, c). b, x a, x
4. The coordinate planes produced indefinitely form eight solid As in plane coordinates the axes Ox angles about the point O. and .cUvicieUtte plane considered into four compartments, so in
O>
space coordinates the planes xy. xz and yz divide the space con-
Compartments four above the plane xy, viz. Q-x y z, Q-x yz and four below it, viz. Q-xyz Q-xy z, O-x y z Q-x yz By an easy extension of the rule of signs laid down in Plane Coordinate Geometry, we regard all x s
slcjttfed
ia.;o
.eight
f
Q-xyz, Q-xy
z,
f
yz as + and on the left of yz as ally s xz as -f- and those behind it as all z s above the plane xy as + and those below it as We can then write the points whose distances from the coordinate planes are a, b and c in
on the
right of the plane
;
in front of the plane
the eight different angles thus
In the
first
Octant, Q-xyz
is
(a, b, c)
In the second Octant, P 2 is In the third Octant, P3 is In the fourth Octant, P 4 In the fifth Octant, P 5 is
is
(a,
( (
b, c)
b,
a,
c]
a, b, c)
c)
(a, b, (a,
(
(
In the sixth Octant, P 6 is In the seventh Octant, I\ is In the eighth Octant, P8 is
b,
b,
c]
a,
c)
c).
a, b,
The
ments.
1.
signs thus
tell
and the lengths of
a, b
and
us in which compartment the point falls, c give us its position in these compart
2.
Construct the points Construct the points
I,
2,
o,
2,
i,
0,0,
3, o,
4, o, 3.
;
i,
3,
3,
2, o, o.
and R are called the projections of P on the 5. The points M, three coordinate planes, and when the axes are rectangular they are its orthogonal will treat mainly of orthogonal pro projections.
We
jections.
we
are to
For shortness sake when we speak simply of projections, be understood to mean orthogonal projections, unless we
state the contrary.
We
which
will give
now some
us.
other properties of orthogonal projections
will
be of use to

6
DEFINITIONS.
is
projection of a line on a plane projections of its points on the plane.
The
the line containing the
When
one
line or several lines connected together enclose a plane
area, the area enclosed
by the projection of the
lines
is
called the
projection of the
first
area.
idea of projection may be in the case of the straight line thus if from the extremities of any limited straight line we draw perpendiculars to a second line, the portion of the latter intercepted between the feet of the perpendiculars is called the projection of the
The
extended:
limited line on the second line.
From this we see that OA, and (coordinates rectangular) are the projections of on the three axes, or the rectangular coordi nates of a point are the projections of its distance from the origin on the
OB
OC
OP
coordinate axes.
7.
I.
FUNDAMENTAL THEOREMS.
The length of
to the line
the projection
of a
finite
right line on any plane
is
equal
tht,
multiplied by the cosine
of
the angle
which
it
makes with
plane.
Let
jection
PQ
;
be the given
finite
straight
line,
;
xOy
draw PM, QN perpendicular to it then tion of PQ on the plane. Now the angle made by
is
MN
the plane of pro is the projec PQ with the plane
the angle
made by PQ
with
MN.
to
MN
= the
PQR.
meeting
angle
PM
made
R, then by PQ with MN.

in
QR = MN,
Now
Through
MN = QR =
Q draw QR parallel and the angle PQR

PQ
cos
(Fig. 2.)
II. The projection on any plane of any bounded plane area is equal to thut area multiplied by the cosine of the angle between the planes.
i.
We
shall
begin with a triangle of which one side
BC
is
parallel
to the plane of projection.
The
area of
ABC = D.
BC
But
x AD, and the
area of the projection
ABC
- BC xA
BC
BC and
A D = AD
planes.
ccs
Hence
ADM. Moreover ADM = the angle between the A B C= ABC x cos angle between the planes. (Fig. 3.)
ABC
is
2. Next take a
rallel to
of which no one of the sides triangle the plane of projection. (Fig. 4.) i*
pa
6
Through

the corner
of the triangle draw
CD
we
parallel to the
plane of projection meeting AB in D. between the planes, then from i C
Now
if
call
= BCD cos = ABC cos
0.
A D = ACD cos 6 A B D -B C D =(ABD - BCD) cos
6 the angle
and B C
or
D ABC
d.
3. Since every polygon may be divided up into a number of it is true also of triangles of each of which the proposition is true
the polygon, i. e., of the sum of the triangles. Also by the theory of limits, curvilinear areas being the limits of polygonal areas, the proposition is also true of these.
8.
The
projection
of a
finite right
line
upon another right
line
is
equal
lines.
to the first line multiplied
by the cosine
of
the angle between the
Let
PQ be
means of
parallel to
the given line and the perpendiculars
MN
it.
its
projection
on the
line CXv,
by
PM
MN
and QN.
Through
is
draw
QR
and equal
to
Then
PQ
PP"
with Ox, and

If there
MN = QR = PQ cos
,
P"
PQR
the angle
made by
PQR.
(Fig. 5.)
9.
on any line will be equal to sum of the projections of PP and P on that line. Let D, D the be the projections of the points P, P on the line AB. between D and Then D will either lie between D and or D In the one case DD" = DD + D and in the other DD"= DD - D"D = in both cases the algebraic sum of DD and D
P",
and P
be three points P, P the projections of
joined by the right lines
PP
PP"
P"
D"
P"
D"
D"
D"
D".
The
is
projection -f or
.
is
-f
or
according as the cosine of the angle above
In general if there be any number of points P, P etc., the pro jection of PP" on any line is equal to the sum of the projections of PP P etc., or, the projection of any one side of a closed po
,
P",
P",
lygonal line on a straight line of the other sides on that line.
is
equal to the
sum
of the projections
10.
USEFUL PARTICULAR CASE.
The projection of th? radius vector of a point P on any line is equal the sum of the projections on that line of tlie coordinates OM, MN, is a closed broken line, and NP of the point P. For
OP
OMNP
the projection of the side on a straight line must be equal to sum of the projections of the sides OM, MN, and NP on that the
line.
OP
AZOTES
11.
OA>
SOLID GEOMETRY.
POINTS.
DISTANCE BETWEEN
Two
Let
z
P and Q, whose
rectangular coordinates are (x,y, z] and (x
y, ), be the two points. (Fig. 6.) We have from the right parallelopipedon
is
the diagonal,
yy Hence
If
2
PQ = PM +
2 2
MN
+ QN
2
.
But
PMNRQ of which PQ PM = x - x, MN
x
PQ =
2 2
NQ= z
(x
2
z.
- xj+
2
(y
-y Y +
(*
- z )\
o,
one of the points P
be at the origin then

.
y = o, 2=0,
and
PQ = * + /
TO
+z"
12.
FIND THE RELATIONS BETWEEN THE COSINES OF THE ANGLES
WHICH A STRAIGHT LINE MAKES WITH THREE RECTANGULAR

AXES.
Take
POjt:
the line
of,
OP
/3,
through the
origin.
{
Let
,
OP
r,
POy
=
x
the angle
POz
r*
y,
and x
2
y
z
the coordinates of P.
Then by
Hence
Art.
8,
1 1,
=
;
,v
+y +
2
-.
But, Art.
= r cos a y = r cos ft r = r (cos a -f cos ft

1
= r cos
i.
y.
-\-
cos 2 ^) or
(i)
cos
2
<?
cos
ft
cos 2
very
im
portant relation.
Cos #, cos
/?,
cos
determine the direction of the line in rectan
gular coordinates, and are hence called the direction cosines of the line. We usually call these cosines /, m and respectively. So the equa
m^ n? tion (i) is usually written P i, (i), and when speak of a line with reference to its direction, we may call
(/,
we wish
it
to
the line
m,
ii).
Only two of the angles
a,
ft,
can be assumed at pleas
ure, for the third, y, will be given
by the equation
cos
2
cos
13.
A/ 1
cos
ft.
We
can use these direction cosines also
for
determining the
position of any plane area with reference to three rectangular coordi For since any two planes make with each other the nate planes.
is made by two lines perpendicular to them respec the angles made by a plane with the rectangular coordinate planes are the angles made by a perpendicular to the plane with the coordinate axes respectively. Thus if be the perpendicular to a
same angle which
tively,
OP
plane, the angle
made by
;
with xz
ft,
the angle ft cos y, are called also the direction cosines of a plane.
is
the plane with the plane xy and with yz is the angle a.
is
the angle
So cos
a, cos
is,
That
the
g
direction cosines
iVOTES
ON SOLID GEOMETRY.
reference to rectangular coordinates are
to this
2
of a plane with of a
line
2
the direction cosines

14.
perpendicular
plane.
i
The
relation cos
a +
cos
ft+ cos
,
y=
enables us to prove
an important property of the orthogonal projections of plane areas. For let A be any plane area, and A x A A z its projections on the
y>
coordinate planes yz, zx and ;ry respectively. A cos a A y A cos ft A a A cos y.
Then
Art. 7, II.,
Az
Squaring and adding we have
A Z2 +
or
A,,
A/+ A/
the square
= A (cos +A =A
A. 9
2 9 2
a +
cos ft
-f
cos y)
That
is,
of any plane area
is
equal
to the
sum of the squares
of its projections on
15.
three planes at right angles to each other.
To
FIND THE COSINE OF THE ANGLES BETWEEN Two LlNES IN cos ft, cos y) TERMS OF THEIR DIRECTION COSINES (cos
<*,
AND
(cos
cos
ft
cos
).
Draw OP, lines. They

and the angle
Let
dinates of
OQ
will
through the origin parallel respectively to the given have the same direction cosines as the given lines,
be the angle between the given lines. (Fig. 7.)
r,
POQ will
}.
POQ = 6, OP = Q (xy z
by Art. n,
(
OQ
coordinates of P(x,y, z\ coor
Now
PQ* =
-x>y
+ (y
-yj +
(Z
- Zj = X* + f
Z*
2}>/
+ X* + / + 2ZZ }.
And from
triangle
POQ,
2 2 r -f r
PQ =
hence
r
9
2rr
cos
z*
0,
2rr cos B
= x* +/+
^
2
+ x*
a
+/
^
z*
2yy
-f
+ 2zz
).
But
Therefore
r*=
.r
+j +
;
and r
,r
+y +
,
2
.
rr cos 6
a cos B
= xx + yy + zz
or
z z x y y = x .+.--+-.. r r r r r r
Hence
cos 6
cos
cos
<*
cos
^cos
#
ft
cos
cos
(i)
which we write
cos B
//
+ ww +
(2)

Cor.
//
i.
If the lines are
+ mm +
Cor.
2.
nri
o (3).
(/,
o or perpendicular to each other cos 6 is called the condition of perpendicu (3)

),
larity
of the two lines
m,
(/
).
From
sin
2
expression
for
cos 6
we
find a
convenient one
for sin
6.
Thus
whence
1
0=1=
(Irti
(//
+ mm +
I
nn
r
=
n
)
2
(/
)
+ m*+
(//
n*} (l
+m*
nn Y
(4)
+
sin 6
2
m}*-}- (ln
(mn
+ mm + m n)\
6.
To
P
express the distance between two points in terms of their
oblique coordinates.
Let
(xyz)
and
(xy z
be the two points.

is
(Fig. 8.)
The
parallelopipedon
v, and the angles made by PQ with the A, //, yOz axes respectively a, ft and y. Project the broken line PMNQ on Hence we will have This projection is equal to TQ itself. PQ.
xOz
MPQN
oblique.
Let
the
angle
xOy
PQ
Now
= PM
cos
a +
MN cos ft + NQ cos
PMNQ
y.
(a]
project the
broken line
on the axes xyz
respectively.
We
obtain thus the three equations
PQ PQ PQ
Now
have
= PM + MN cos + NQ cos = PM cos A + MN + NQ cos v ft cos y PM cos /* + MN cos v + NQ

cos
A.
/*
cos
V
)
(b)
multiply the
first
and the third by

2
NQ
of equations (3) by PM, the second by and add them taking (a) into account and we
MN
PQ
or
= PM + MN
2
+ NQ + 2PM
2
MN cos A
/*
+ 2 PM NQ + 2MN,NQcos v
.
cos
(c)
)
PQ
Cor.
=
If
2
- XJ +
2(x
(y
-/) +
z
)
- z Y+ (z
+ 2(y
-y
2(x
)(z
x
)
(y
-y
cos
/I
x )(z
cosfii
cos v.
(5)
one of the points as

x* + y*
Q be
at the origin then
PO
17.
z* -f
2xy cos
A+
2x2 cos
jn
2zy cos v.
(6)
Direction Ratios.
In oblique coordinates the position of a line
PQ
call
is
determined by the
PM
ratios
MN
;
NQ
;
-j-y
-^
these
~p^-,
m,
and these we
direction ratios.
We may name
/,
n respectively,
IO

call
taking care to note that \ve are using oblique coordinates and
the line
PQ, the
line
(/,
;//,
n}.
To
find a relation
16,
among
2
.
these
direction ratios,
we divide equation
2 ;/
(c) Art.
by
PQ
We
(7)
thus
have
i
m*
2//T2
cos
2ln cos
/*
+ 2mn
cos v,
the
desired relation.
1 8. The coordinates of the point (xyz) dividing in the ration the distances between the two points (x y z ) are
1 :
(x"y"z")
mx"
=.
+ nx m + n
y J
my"
7?i
+ +
ny n
z= -
mz"
+ m +
nz
.
(8)
The proof of this is precisely the same as that for the correspond ing theorem in Plane Coordinate Geometry.
19.
POLAR COORDINATES.
position of a point in space is also sometimes expressed by the following polar coordinates 6 which the radius The radius vector r, the angle PO0
:
The
OP =
vector
makes with a
projection
OC
the and the angle CO^tr cp which of the radius vector on a p ane yQx perpendicular to
fixed axis Oz,
O0 makes
with the fixed line
Ox
in that plane.
the formulae for transforming from rectangular to these polar coordinates are
6.
We
have
OC =
(Fig. 9.)
r sin
Hence
x
Y z
r sin 8 cos
<p
r sin 8 sin
<p\
(9)
r cos 8
2
and these give
r1
x^ +.V
tan cp
=
(10).
a cos 8
a Conceive a sphere described from the centre O, with a radius and let this represent the earth. Then, if the plane zOx be the plane of the first meridian and the axis of z the axis of the earth,
Q
face.
latitude,
cp
longitude of a point on the earth
sur

20.
Ir
Distance between two points in space in polar coordinates.

r
,
MN
OQ
Let P be (r
is
q>
and
(r, 6,
<p).
Project
PQ on
the plane xy,
this projection,
draw
OM
and
ON
the projections of
OP
and
respectively
then
PR
= MN.
have
on that plane.
(Fig. 10.)
Through P draw PR
parallel to
MN,
And we
But
PQ = PR + RQ
2 2
= MN +
2
.
(QN
RN)
2
.
in triangle
2
MON MN == OM + ON - 2OM ON cos MON,

8
or
MN = r
2
sin
&+
2
r sin
2rr sin
sin
d cos
r cos 6
sin
(tp
.

cos^
+ sin
6 cos
(
)).
(n)
CHAPTER
II.
INTERPRTTATION OF EQUATIONS.
TRIPLANAR COORDINATES.
21. LET us take F (x y, z] o, that is any single equation con and z. This may be considered as a taining three variables x, relation which enables us to determine any one of the variables when
y
the other two are given. may be written
Let these be
x and j
So the equation
*=/(*, jOi
in
y.
which we may attribute arbitrary and independent values to x and And to every pair of such values there is a determinate point in the
;
plane xy
and
if
through each of these points we draw a line

it it
parallel
to the axis of
z,
and take on
is
by the equation,
lengths equal to the values of z given clear that in this way we will get a series of
points the locus of which is a surface, .and not a solid since determinate lengths on each of the lines drawn parallel to z.
we take Hence
(x, y, z)
22.
=o
represents a surface in triplanar coordinates.
If the equation contains only
two variables as
(x,y)
=o
then
it
represents a cylindrical surface.
For F (x, y] o is satisfied by certain values of x and y inde pendently of 0, and x and y are no longer arbitrary but one is given in terms of the other to each pair of values corresponds a point in
;
and the locus of these points is a curve in that plane. If through each point in this curve we draw a coordinate parallel to 2, every point in that coordinate has the same coordinates x andj/ as the point in which it meets the plane xy. Hence F (x,y) = o repre sents a surface which is the locus of straight lines drawn through points of the curve F(.r, v) = o in the plane xy and parallel to the
the plane xy,
12

axis of
z.
13
called a cylindrical surface with axis parallel to z or a plane parallel to the axis of z according o in the plane xy represents a curve or a as the equation F (x,y)
This locus
is
either
what
is
straight line.
1 For example, x
2 r o in rectangular coordinates is a y* a r* is a wiih circular base in plane xy (since jv right cylinder circle in plane xy) and its axis coincident with the axis of z.
4-
+y =
c o is a plane parallel to the axis of g, intersect And ax + by c. ing the plane xy in the line ax + by o represents either a cylindrical surface with Similarly F (x, z)
. axis parallel tojy or a plane parallel o represents either a cylindrical surface with axis parallel F (y, z)
to_>
to the axis of
23.
or a plane parallel to this axis.
equation containing a single variable represents a plane or planes parallel to one of the coordinate planes. a represents a plane parallel to the planejyz. Thus x
An
And
as _/"(.#)
values of x, as
= o when solved will x= x = a, x

I,
give a determinate
c,
number of
etc.,
so
it
represents several
planes parallel to the coordinate planers. o represents a number of planes parallel to the Thus also F(jy) plane xz.
And F
24.
(z)
= o,
number
of planes parallel to xy.
Thus we
it
see that in all cases
preted
represents a surface of
when a single equation some kind or other.
is
inter
The apparent exceptions to from their nature can only be

must
(x
satisfied
this are those single

satisfied
satisfied.
z
equations which
exist
simultaneously are
b)*
when several equations which As for example

equation can only be
z
of + (y
when (x
b,
(z
a)*
= o,
c)
(y
= o. This = o, (z b}*
c}
o,
or
x=
a,
c.
Now
these represent three

a,
b,
c.
planes, but being simultaneous they
= o is only satisfied by x = a, y = b, -f ( y b}* and hence though x = a is a plane, and y = b is a plane, the two together must represent a line common to both of these planes, that is their line of interseciion, which must be parallel to z.
25.
represent the point z So also (x a)
In general two simultaneous equations as
f(*,y,
*)
F (x,y,
z)
=o
14
represent a curve or curves, the intersections of the two surfaces represented by the two equations.
Thus
taken simultaneously we have seen represent a straight

z,
line parallel to the axis of
F (x)
=o
_
the intersection of these two planes.
p)
axis of
represent a
number
of straight lines parallel to the
z,
the intersections of the several planes parallel respectively to
the planes yz
and
)
xz.
F
of y,
(x)
=o
p/j)
e;c.
of
re P resent a
numbe r
of straight lines parallel to the axis
(.v,
z)
=
y)
f.
j
represent the curves of intersection of the two cylin*
ders
(x,
= o and
(x,
z)
= o,
e c., etc.
26.
Three simultaneous equations
F (x, y, z}=o\
as
f(x, y,
z)
=o
or
F F
(x,
(A-,
j>)
z)
= o\ = o I etc
.,
represent points in space or the intersections of the lines of intersec tion of the surfaces.
The
simplest case
is,
y
z
^
>
representing the point
(a, b, c],
c]
So also
2z
\
x +y
2z
v.
represent points which
can be found by
solving the three equations which themselves represent different sur

faces.
Interpretation
27.
of Polar Equations.
i.
r
Hence
r
represents a sphere having the pole for its centre. o which gives values for r as r the equation a, (r)
r
=a
b,
c,
etc.,
represents a series of concentric spheres about the
pole as centre.

2. 6
jcj
a represents a cone of revolution about the axis of z with vertex at the origin of which the vertical angle is equal to 2a. o giving values 6 Hence the equation F (6) a, 6 ft, etc.,
its
axis of z having the origin for represents a series of cones about the a common vertex.
a plane containing the axis of z whose line (3 represents 3. cp of intersection with the plane xy makes an angle /3 with the axis of o which gives values q) x. Hence the equation F (cp) (3, cp
=
F
of
ft
etc.,
represents several planes containing the axis of z inclined

,
to the plane
4.
(r,
cp,
zOx at angles ft, ft etc. the equation involve only rand 6 as F (r, 6) o, since the same relation between r and 6 for any value o gives 6) ii gives the same curve in any one of the planes determined by
If
Hence it represents a surface of revolution assigning values to (p. traced by this curve revolving about the axis of z.
xz, or in
a cos 6 is the equation of a circle in the plane a cos 6 Hence r plane containing the axis of z. any represen s a sphere described by revolving this circle about the axis
Example,
of*.
5.
If the
equation be
are one or
more
the po e
containing
may Oz
o for every value of cp there F(<p, 0) values of 6 corresponding to which lines through be drawn, and as cp changes or the plane fixed by it
revolves,
these
lines
take
new
(a
plane, position conical surface being any surface generated by a straight line moving in any manner about a fixed straight line which it inter
sects.
)
of the
and
thus
positions in each generate conical surfaces
new
6.
If the
equation be F(r,
(p)
= o, for
every value of (p there are
one or more values of r, thus giving several concentric circles about As (p the pole in the plane determined by the assigned value of cp.
changes, or the plane through
Oz
revolves these values of r change,
and the concentric
circles vary in
magnitude.
The equation
thus
represents a surface generated by circles having their centres at the pole, which vary in magnitude as their planes revolve about the axis
of z which they all contain. 7. If the equation be F

general.
o, it represents a surface in (r, 6, cp) F (r, 8, ft) as cp assign a value to ft, then o will represent a curve in the plane (p And as cp changes ft. or the plane revolves about Oz this curve changes, and the equation
For
if
we
<p
will represent the surface
containing
all
these curves.
!6
28.

Two
simultaneous equations in polar coordinates represent a And three simulta the intersections of two surfaces.
the intersections of three
line, or lines
neous equations represent a point or points

surfaces.
Thus
r
=a
a
j
9 = ft)
taken simultaneously represent points determined
by the intersection of a sphere, cone and plane.
CHAPTER
III.
EQUATION OF A PLANE.
COORDINATES OBLIQUE OR RECTANGULAR.
29.
To find equation of a plane
in
terms of the perpendicular from the
origin
and
its
direction cosines.
Let
OD
and
let it
p be the perpendicular from the origin on the plane, make with the axes O,r, Oy and Qz the angles a, ft and y
Let
respectively.
OP
be the radius vector of any point

(Fig.
is
of the
plane
The
OM, MN and NP the coordinates of P. projection of OM + MN + NP on OD
n.)
equal to the pro
jection of
OP
on OD.
The
projection of
OP
x
on
is
OD
x a
is
OD
itself,
and the projection of

ft
OM + MN + NP
Hence we have
30.
on
OD
cos
a +y
cos ft
in
cos
z cos y.
cos
+y
z cos
p.
(12)
To find
the equation
of a plane
terms of its intercepts on the
coordinate axes (coordinates oblique or rectangular).
Let the intercepts be (12) may be written
OA
y
a,
OB =
b,
OC =
c.
The equation
x
p
But since
sec
sec ft
sec
y
have
sec
= OA
ODA, ODB and ODC are right-angled triangles, we = b, p sec y = OC = c. a, p sec ft = OB
Therefore the equation becomes
the equation of the plane in terms of
its
intercepts.
8
31.

Any
equation
Ax + By + Cz
plane*
=D
(14) of the first degree in
x,
y and z is the equation of a For we may write (14)
j^
J>/
"B"
_z_
Jl
X
And
Hence
_D/
~C~
D
putting
is
D
-g-
-^= a,
D = ^ --= A
We
have
the
form
(13).
(14) coordinates.
the equation of a plane in oblique or rectangular
Hence
to find the intercepts of a
the coordinate axes,

raakej>
o and z
=
;
find intercept onj/

"Example.
x=
plane given by its equation on we either put it in the form (13) or simply o to find intercept on x z = o and x = o to o andjy o to find intercept on z.
;
Find the intercepts of the plane 2x
$y
52
=
Cz
60.
32.
It is useful often to
reduce the equation AJC

ft
By
=D
to the nates.
form
cos
a +y
cos
z cos
=/
in rectangular coordi
We
derive a rule for
this.
Since both of these equations are to represent the same plane, we have
cos
_ cos (3 _ cos y _ p _ Vcos + cos (3 + cos A B D yT7 A + B + C C

2
a.
""
Hence
cos
<*
=
+ B +C
2
-v/A
+B +C
2
+ B +C
2 2 ! A/A + B +C
=
it is
-vWWc*
(I5)
in the perpendicular
form (12).

Hence
the
ig
each term of the equation Ax + By sum of the squares of the coefficients of ofx, y and z, the new coefficients will be the direction cosines of the per pendicular to the plane from the origin, and the absolute term will be the Give the radical the sign of D. length of this perpendicular.
= D,
Rule: If we
divide
by the square root
the
Find the direction cosines of the plane 2x Example. 6 and the length of the perpendicular from the origin.
Result.
$y
40
cos
=
A/4
===
+
16
7=,
29
cos
ft
T=,
2
cos
=,
4
29
P
33.
lar).
V 29
To find
the
angle between two planes
(coordinates
rectangu
If the planes are in the
form
y cos cos
ft
ft
x x
cos cos
is
a +
z cos
=p
p
,
a +y
z cos
then since this angle
origin on the planes the cosine
equal to the angle of two perpendiculars from will be (Art. 15) cos cos a cos a
V=
cos
/3
cos
/3
-f
cos
cos
If they are in the
form
A.v
+ By +
B>
A .v +
Then
cos
=D + Cz = D
Cs
cos
/?
==.
B2
4-
C2
r,
=
A/A
cos
2
+ B + C
2
=
B
2
COS
flf
A
~,
COS
al
=
A/A
+B +
/f
cos
=
2
And
cos
AA
AA + BB + CC + B + C VA + B + C
2 /f
2Q

this
From
.
_
d
.
(A
-f
B2 + C
)
2
(A
+ B +C + B* + C )(A
2 2
(A
)
2
-(AA + BB + CC
2
+B
2
2
C"
v-
(AB
-A B)
2
+(AC
-A C)
8
(A
Cor.
. .
+ B + C ) (A
is
+(BC + B +C
a
-B C)
2
i.
If the planes are perpendicular to each other, then cos
V=o.
the
AA + BB + CC = o
Cor.
2.
(18)
the condition
of perpendicularity of
o.
planes.
If the planes are parallel sin
V=
2
Hence
2
or
+ (AC - A C) + (BC - B C) = o AB - A B = o AC - A C = o BC - B C = o
(AB
- A B)
ABCC
A7== B 7=
32
or the condition that the two planes shall be parallel, is that the and z in the two equations shall be proportional.
coefficients
ofx,y
Ex.
i.
Find the angle between the planes
x+
2.
2y
and $x
4y
4- z
10.
Show
that the planes
x + 3y
dicular to each other.
3.
5Z
20 anc* 2X
I0 are perpen
Write the equation representing planes parallel to the plane 3*

6z
2y
34.
ii.
To find
the expression for the distance from a point
(x y z
to
plane (coordinates rectangular).
i. Let the equation of the plane be of the form
cos
a +y
in
cos
ft
z cos
= / when p
OD.
and produce
Pass a plane through
parallel to the given plane,
OD
to
meet
it
The equation
of this plane will be
cos
+y cos
ft
cos
=p
when
OD =/
Now
Then
let
PM be the perpendicular from P PM = OD - OD = p - p.

1
on the given plane.

Hence
21
PM = x
(x cos
1
And
for the
cos a 4- y cos a + y cos /? +

the sign
/3
z cos
p.
is
cos
p) (20)
the expression
perpendicular from the point
cos
ft
cos
y =p,
xy z on being + or
the plane
cos
or
is
+ _y
or
is
according as
not on the side of the plane remote from the origin.
2. Let
the plane be in the form
Kx +
By
Cz
D.
Then
cos
V A + B + C*
2
- etc., etc.
(15) Art. 32.
Hence
the expression
(y
cos
-FJF
cos
/3
z cos
p] becomes
-D
Ex. Find the length of perpendicular from the point
the plane
3Jt-
(3, 2,
i)
on
4_y
6z
6
=
24
-
24.
Result.
9 4- 8
A/9+ 16+
13
36
35.
cos
=p
The equation of the plane in the form x cos a + y cos (3 4may be used to demonstrate the following theorem in
for
its
projec ions. 7% volume of the tetrahedron which has the origin

the triangle
vertex
and
ABCy2?r
its
base
is
equal
to the three
pyramids which have any
point (x, y, z) in the plane

the projections
respectively.
of the area
ABC for their common vertex and for bases ABC on the three rectangular coordinate planes
For
let
be the area of the triangle
ABC
z cos
and
x
the equation of
its
cos
+y cos
/3
=p
=
plane.
Multiply
this
equation by A.
Then
or
cos
a x
.
.
-f
cos
/3 ft
y+ A
.y
cos
Ap
-JA cos
a x + Â
cos
-JA cos y. s
But A cos a, cos /3, A cos y, are the projections of on the planes yz, xz, and xy respectively, and x, y and z are the altitudes of the tetrahedrons which have these projections as bases and the point is the volume of the pyramid (x, y, z) as common vertex, and
-JA/>
22

for vertex
which has the origin

is
and
for base.
Hence
the theorem
true.
Calling these projections

tion of the plane
36.
A^v
+A
Ax Ay + A^
,
= 3V.
.
and A x we may write the equa

,
(22)
To find
the polar equation
of a plane
cp
Let
OP =
r,
POS
6,
OM =
be the polar coordinates of a

angle
(Fig. 12.) point P of the plane. a Let perpendicular on plane
OD =
DOS
a,
D OM
order to
jff,
and
POD =
GO.
Then yyp -
cos
POD =
cos
GO,
or
cos
GO.
Now in
express GJ in polar coordinates conceive a sphere about O as centre Draw r as radius. with OP to D" on the sphere. Prolong
the arcs of great circles SPP , The triangle SD"P has for
SD"D
OD MP D
,
and
D"P.
its
sides
<p.
SD"
a,
SP SP
(fi
=
cos
6,
D"P
GJ
and angle
cos
D"SP
= D OP = /3
SD"
But
D"P
cos
cos
SP
+
sin
sin
SD"
sin
D"SP.
Or
cos GO
cos
cos 6
sin
6 cos
cp).
Therefore
cos
cos 6
sin
sin 6 cos (f3
cp)
(23)
is
the
polar equation of the plane.

37. The general equation of the plane be reduced to the form
Ax + By
4-
Cz
=D
may
Ax
And
z
4 By
f
+ C2=
c
(24) by dividing by the absolute term D.
also to the form
mx + ny+
(25)
by dividing by
transposing and putting

in
= m,
V^/
=
V^
n and -^
v_/
c.
These two forms are very useful
the solution of problems
and
in finding the
equations of the plane
under given conditions.
Plane tinder Given Conditions.
general form
i. The equation of a plane through the origin will be of the Ax + By + Cz = o, for ihe equation must be satisfied and 2 = 0. by ^ = 0,^ = 2. The equation of a plane which contains the axis of z is of the
38.
NOTES
form
SOLID GEOMETRY,
is
Ax +
;
By = o a plane containing the axis of y o. one containing the axis of x is By + Cz

;
Ax +
Gar
a plane parallel to the axis of z is Ax + By one parallel of one parallel to the axis ofjy is AJI* + Cz D; D. to the axis of x is By + Cz
3. The equation of
=D
4. The equation
parallel to
^0
is
By = D
of a plane parallel to the plane j/s

;
is A^t:
=D
parallel to
xy
is
Cs
D.
These equations we have had already

To find
to
in the forms
x=
a,
3,
and parallel
of a plane containing a given point (a, b, c) i a given plane Ax -f By + Cz D. ( ) First, since the required plane is to be parallel to (i ) it may be writ is undetermined. ten A* + Bj/ + Secondly, the (2) when
39.
the equation
G&=D
coordinates
(a, b, c)
must
satisfy (2).
Therefore
Aa + ~B& + Cc
=D
Hence by subtraction we eliminate
and obtain
Ax +
By
C0
Aa
B3
(26)
the required equation.
Example.
point
40.
(i,
2,
Find the equation of the plane passing through the 6. 4) parallel to the plane 2x + 4y 32
To find
,
the equation
y",
(x
),
(x",
z")
of a plane passing through and (x z y

",
",
three given points
").
A,
Let the equation of the plane be of the form Ax + By B and C to be determined by the given conditions.
Since the plane
is
Cz
i,
to contain
each of the points, we must have
Ax + B
Hence
X,
JP",
I,
^
z 2
i,
i,
I,/",
B=
",
C=
Substituting
these
values in
the
equation
Ax + By + Cz

we have
#,
I, I,
,
I,/
*,
x+
r
x"
*
(27)
i,
But from plane coordinate geometry the coefficients of x, y and z in these equations are the double areas of triangles in the planes jy0, xz and xy respectively. Moreover these triangles are the projections
of the triangle of the three given points, on these planes. comparing this equation wi h the equation (22)
,z
Hence
= =
3V
6 times the
we
see that
x ,y
,ss
= 6V.
That
is
volume of the
x">y",
given points for base.
pyramid which has the origin for vertex and the triangle of the three This equation fully written out is
^(yv--yv )+^ (yv-y3 )+y

41. To find the equation
(jV -yv)-6v,
contain the line
2 8)
tion
+ By + Cz = D and Ax 4- By + Cz = D + C 0-D )=o(29) ThisequationisA# + By + Gs D + K(A jc +

of the two planes Ax
.
of the planes which
of intersec
/
B>
when
particular value and
K
is
plane when K takes a D = o and when A^ + By + C-s: D = o are satisfied simultaneously. Hence it x 4- B^ + C z Hence as K is arbi a plane containing their line of intersection.
is
arbitrary.
For
this represents a
it
is satisfied
trary
it
(24) represents the planes containing the line of intersection
of the two given planes.

42.
the
When
each
the identity
equations
U=o, U
KU + KÛj + KaUs = o (30) exists between o, U = o of three planes, then these planes
3
intersect
other in one
and
the
same straight
line.
This
is
an
Also when the equation of the first easy corollary of Article 41. degree in x, y and z contains a single arbitrary constant all the
planes which
line
it
stant intersect each other in
expresses by assigning particular values to this con one and the same straight line. This
of intersection
may be
at infinity
and then the planes are
all
parallel.
Example
=. 3
I.
The
(M arbitrary) 3 planes 6x+ 22
planes represented by the equation 6x+T&y + 2z all contain the line of intersection of the two
and
o.

Example
2.
25
42
The
planes
represented
by 2x
+ ^y
(n
arbitrary) are parallel.
Example.
The
planes
$x
4.v
+ X+
\v
2V
4T
+ + +
60
3
=
=
2 }
I
&z
2 J
intersect in
one and the same
straight line because

2
=o
o,
is
an
43.
identity.
Wfo
between the equations

o,
1
of four planes
in
any form
U,
U o the identity KU + K U + K U -hK U = o

o,
2 3
1
(31) ^/j/j, then
intersect each
other in one
and
first
the
sam*
point.
o,
dinates which satisfy the o. 3 satisfy the fourth
three
U=
these four planes For then any coor U, = o and U 2 = o will
U =
44.
the origin
Ax +
i. Find the equation of the plane passing through and containing the line of intersection of the two planes + C z = i. By + C0 = i and A x +
Example
B>
First
we have
Ax +
By
Cz
+ K (A .r +
B>
+ Cz
1)= o
for all the
planes.
planes containing the line of intersection of the two given But as the required plane must contain the origin, the
satisfied
equation must be
.-.
K=-
by
(o, o, o).
Hence we have
K=o.
i.
The
required equation
is
therefore
Ax +
or
By
-f
Cz
)
(A x
4-
B>
+ Cz
i)
=
=
o
o.
(A
Ex.
2.
- A x
(B
B )^ + (C
On the three axes of x, y and z take OA = a, OB b, and construct on these a parallelopipedon having MP as the edge opposite parallel to OC, and AR in the plane xz the edge opposite and parallel to BN. Find the equation to the plane containing the three points M, N and R.
OC =
Now NR
\?
is
the line of intersection of the two planes
+-7b
and
-=
i.
Hence
the plane containing this line
must be of
the form
26
OC
A
V
"*~
OTES ON SOLID GEOMETRY.
-I
l )
To
determine
\ve
impose the
b, o).
condition that this plane shall pass through the po nt
Hence we have
+ at
v
b
-r
+K
is
\c
z
= o.
M (a, K=
i
Therefore the required plane
x
a
--t-V
in like
H---c
o or
x v z --(-^-_j-__
a
b
c
2.
Ex.
3.
Find
manner
the equation of the plane containing
the points P, B,
and
C, in the
same
figure.
Result,
45.
If
two given planes be

cos
ft
in the
normal form as
x
x
cos
+>
+z +z
i
cos
y=p
y
and.v cos
a +y
cos
is
/3
+ zcos y = p
+z
cos
The plane containing

cos
their line of intersection
a+y
cos
/3
cos
p-\-K (x cos
a +y
cos
ft
And if K = x cos a -\-y cos
-/)=o
the equation
becomes
(x cos
/3
+z
cos
a +y
cos
ft
+z
cos
;/
-/>
)=
which represents ihe two plane bisectors of the supplementary angles
made by
the given planes.

bisectors
That is to find the equations to the plane angles made by two given planes, put their and then add and subtract them.
Example.
angles
of
the supplementary
equations in the
normal form
made by
Find the two planes which bisect the supplementary the planes 2.v + 3.y Vz = 5 and 3^ + 4^23 =-- 4.
Result,
..
A = x cos + cos y
V29 (Y + y crs
.
/>
=
ft
A/14
Remark.
If
we
A =x
Then
cos
a +
place v cos /?
z cos
and
A=
tween the planes
the plane bisector of one of the angles be o is that of the supple and A and A -f A
is
mentary angle. 46. The three planes which bisect the diedral angles of a triedral have. = o be three Let A = o, A = o and a common line of intersection.
A"

planes in the normal form, and
let
the origin be within the triedral angle
formed by the three of which P their point of intersection is the vertex. Then the plane bisectors of the angles made by these planes is And as these when added A o, A o. A A o,
A" A"
together vanish simultaneously, it follows that these three have a common line of intersection.
planes
We can give this theorem another form by conceiving a sphere to be described about the vertex of the triangular pyramid as a centre. The
three planes o, A"=o cut the surface of the sphere in o, A arcs of great circles which form a spherical triangle and the three
A=
planes
A =o,
their
A"
o and
A"
=o
cut the sphere in
three arcs of great circles which bisect the angles of this spherical
triangle
and
common
line of intersection pierces the sphere in
the
common
intersection of these arcs.
Hence
the above
demon
which
strates the following

bisect the
theorem, namely, The arcs of great
circles
angles of a spherical triangle cut each other
in the
same point
(the pole of the inscribed circle of the triangle).

47.
To find
the
point of intersection of the planes AJC
By
Cs
D,
Ax +
D,
B,
,
B>
+ Cz
=D
A",r
+ B
C C
C"
>
C"z
D".
We
have by elimination
D
A,
C C
C"
A,
D,
,
A,
B,
,
D
B,
D D
D"
D".
B",
A",
D",
A",
B",
B,
,
C C
C"
[A,
C C
C"
A,
B,
,
C C
C"
(32)
A",
B",
A",
B",
A",
B",
Hence
the condition that
one of these
shall be parallel to the line
of intersection of the other two, or that the planes shall not a point, is
meet
in
A,
B,
,
C
,
A
A(B C"-B"C
47.
C
C"
A",
B",
o,
that
is
+ A
(B"C
BC")
A"(BC
B C)
= o.
The condition
that four planes

1
Ax -fBy +Cz +D = o Ax + + Cz + D =o
B>
A"x
+B
>
+C"z
+D"
=o
=
o
shall
"x+B
"y
+C
"z
+ ~D
"
meet
in a point
is
28

A,
B,
,
C,
,
A
A
|
B B
D D
D"
A",
B",
C",
",
",
C",
"
o.
(33)
49.
We have seen D = o and A. x+
that the equations of
two planes Ajr
Bjy
+ Cz
D=
o added together one or both
of them multiplied by any
number
give the equation of a plane
which contains the
we combine
If two given planes. these two equations so as to eliminate x we shall obtain a plane parallel to the axis of A\ containing this line of intersection. If we eliminate y we obtain a plane parallel to the axis of y contain
line of intersection of the
ing the same line
and
finally if
parallel to the axis of z containing the
we eliminate z we obtain same line.
a plane
CHAPTER
IV.
THE STRAIGHT
50.
LINE.
The equations of any two planes taken simultaneously represent
their line
of intersection.
re P resent a strai S ht
lme the
ordinates of every point of which will satisfy the two equations. If we eliminate alternately x and y between these equations
we
obtain equations of the form
x m p\ y ~ nz + q axes Oy and Ox
\
planes
parallel
respectively
to
the
(35).
which represent the same straight line as equations These non-symmetrical forms (35) are very useful. The
=.
planes
line
mz + a, y
= nz + b
are called the projecting planes of the
on the planes of xz and jar, and these equations are also the equations of the projections of the line on those planes respectively.
If
we eliminate
we
get
~= -m
or J y
x+g
-- p the equam
tion of the projection of the line
on the coordinate plane xy.
The equations
stants,
(35) of the straight line contain four arbitrary con
to which we can give proper significance by com mx + b in plane coor paring these equations with the equation y dinate geometry. The equations (35) may be thrown in the form
m,
n, p, q,
^_^ =z _^ = ,
m
n
i
which gives us an easy choice of fixing the of any two of its projecting planes.
51.
line
by the equations
To find the equations of a straight line in terms of its direction cosines
and
the coordinates a, b, c
of a point on
the line
(axis rectangular.}
29
30
Let a,
ft,
respectively.
(x, y, z)
/
Let
be the angles made by the line with the coordinate axes / be the portion of the line between any point
line
;
on the
cos
ft
yb
cos
and the point (a, b, c). Then / cos a = and eliminating / we have y
a-
zc
cos
ex
cos
ft
cosy
(37)
line
This form (37) of the equation of a straight
is
symmetrical
and
It contains six constants but in therefore very useful. reality four independent constants, since the relation cos 2 o cos 2 ft only 2 i 4- cos y holds, and of the three a, b, c one of them may be
is
assumed
We
(37).
at will, leaving only two independent. have seen that the equation (35) may be thrown into the form So also (37) may be thrown into the form (35) by finding
from them expressions

52.
forjy
and x
in
terms of
z.
To find the
direction cosines
of any
straight line given by its equations.
If the equations be in the
form
T
x
=
vbzc
r^
HT
L,
and
XT N
are proportional
to the
direction cosines of the line.
So that we have
cos
_ cos ft _ cos y __ -\Xcos- a + cos ~

~M~~
ft
cos 2
Hence
Hence
to find the direction cosines of any straight line
A.r
MX
we throw
+ By + Cz = D 4+ Cz=D
B>
the equations into theform
.va _ yb _ z c ~N~ L
"IT"
cosines as above by eliminating y and x, and. then write out the direction to each denominator divided by the square root of the sum of the equal
squares of all three.

Thus
to find the direction cosines of the line
nz
we
X
write
it
V
"
m m
cos
Hence
=^+*+r. cos
lines
*>
(39)
Ex.
i.
Find the direction cosines of the
y~
,,
_
.
>
.,
3^-1
\
j
>
3x
y+2z
= -
24 10
U;
given by the equa
53.
tions
To find
the cosine
of the angle between two
lines
xa yb
L
M
cos
and
x
ri
= vb M
:
zc
N
^
~
We
have shown (Art. 15)

cos
V=
cos
cos
-f
cos
/3
cos
/?
cos
cos
y.
Hence
V=
LL + MM + NN *+ N VL *+
s
M"
+N
TT.U iu If the lines be
m
V
the
.\
form
mz-\-p] nz + q
= mz-\-p = n z+q
.
\
\
Then
cos
wm+nn +i 2/
^=
=..
2
,
(41)
Ex.
i.
Find the cosine of the angle between the lines
X**
2Z
+6
lines
Ex.
2.
Find the cosine of the angle between the
xy
=4)
x+y + z
=2p
32

These equations may be put
in the
forms
=
.
= 4-3
_
_
(I
)
and
___ =
-5
cos
V=
,T
-+
5
12
V26X38
54.
The condition
of perpendicularity of two lines given by the
equations in
last article is
LL + MM + NN ^
2
o.
(42)
The
is
condition that they shall be parallel (see Art. 15)

/
(LM
1
A
-L
IV J.
/ 2 M) + (LN -L N)
+(MN
-M N)
i=o
lines are in the
or
-=^=-(43).
IN
These two conditions when the
x
forms
y
= mz+p
nz
= m z+p
n z+q
+q
become
tively.
mm + nn +i
=. o, (44)
and
==
ri
(45) respec
55.
To find
the condition
of the
}
intersection
of two
lines
x
V
= m,+p
nz
+q
This
derived by eliminating x, y and z from the four equations. (m m }z +pf> Subtracting the third from the first we have o
is
/0
_P~P
z Similarly from the second and fourth
~~
n r^_ n
,
and since the

fore
lines intersect these
we have
-^^= m m
/
^
n
,
two values of
(46)
z are equal.
There
~
Ex. Find
tersect.
so that the lines
If the
two
lines are in the
form
xa
~TT
yb
TT~
zc
"N~
(I)
a __ y =
~T7"
~W~
_z
N
r

the elimination of x,
33
readily by writ
and
.
z can be effected
more
ing
(i)=
K
.-.
and
(2)
=K
x-a
x
a
^LK
)
.
<I
= r Vif LK
j
=I*K
LK.
Similarly b- b
MK
r
MK
f-^
Therefore eliminating
^ N K -NK.
K
L,
and
K
,
we have
M,
N,
(
-M
-N
-L a-a , ,
r-/ =ro
or
(47)
The Straight Line under Given
Conditions.
56. The equations of a straight line parallel to one of the coordi nate planes as xy are z mx -\-p. v c, The equations of a straight line parallel to one of the coordinate
axes as
x=
z.
are
y
5 7.
7 o find the equations of a straight line, passing through a given point. y,

z
)
If
(x,
is
the point
is
we have seen
or
,
the equation
xx =-vy
==
= z -^ z
,
,
ft
(4)
if
the equations are in
the form
x = mz+p) * \ y = nz + q
j
then
xx
^r-y
m(zz\}
} = n(zz
,/
)
TT
\ (49). v v
j
Hence
.,.
if
the
equations of a straight line contain only two arbitrary constants, all the lines obtained by assigning values to these arbitraries pass through
a single point.
58.
To find
,
the equations
of a
straight line passing through two given
points (x
(x",
y",
z")
using (48)
we have
~
.L
IVl
"
=
JN
or dividing (48) by this to eliminate
3*
34
L,

M, and
we have
yy zz :?r=P=-77=7wIf
xx
<5>
one of the points as
(x",y,
z")
be at the origin then the equa
tions
become
59.
To find
the equation
}
point (x ,y, z
of a straight line passing through a given and parallel to a given straight line
xa
L
From
,
yb
M
the
first
condition we must have
-W~ ~ -L
.
-,
and Irom the second condition
-=r
L M N =-r T -.=-:. Tr LT M N
1
Hence
the required equation
is
x x L
If the
__
yy _ zz
M
~~N
w2
equation passing through the point

)
\
A--JV
vy
= m (z-z ~ n (z z /!
//
)
)
an d the given line be
x ,y z be x == mz+p
,
of the form
nz + q.
Then
n and
/,
and the
line will be
^2$$Y
60.
To find
r
,
the equations
point
j
y
z b
and perpendicular
of a straight line passing through a given to and intersecting a given right line
x
I
_z
n
The
required line by the
first
condition will be of the form
where L, M, and
are to be determined by the conditions
o (Art. 54)

and
61.
Ex.
c,
i.
Find the equation of the line joining the points
(b, c,
a)
and
(a,
It)
and show
that
it is
perpendicular to the line joining the

;
origin
is
and the point midway between these two points
and
c
that
it
also perpendicular to the lines
=y =
and
a
=-= -. b
which join the middle points of the Ex. straight sides of a tetrahedron all pass through one point. opposite Take O one of the vertices as origin and OA, OB, OC as the
2.
The
lines
axes of x, j
s.
M"
be the middle points of BC, AC and respec the middle points of the edges OA, and AB Then to find the equations of the opposite to those respectively.
Let M,
N,
tively,
N"
OC OB
lines
MN,
MN
M"N".
We
(M,
apply the equation
x
2b
-=
y y
2c. o).
,,
-77-
to
the points
N) (M
N
20,
;
(M",
N")
respectively.
Let
OA =
OB =
c)
OC =
is
Then
is (o,
b,
and
(a,
o,
Hence
the equation of
J
~
MN
(l
is
^~^
a~~
Similarly the equation of
_.^_ b
~^
c
is
/
T
MN
b
x
a
__yb _z
c
is
And
the equation of
M"N"
(i)
and
(2) give
=, z =
and these values

,
satisfy (3).
Consequently these lines pass through the point (-,
-).
36
A 07^S OA SOLID GEOMETRY.
Straight Line, and Plane.

62.
To find
its
the conditions that a line shall be perpendicular to
a plane
equation. given by If the plane be of the form
.v
cos
we know
that cos
<v,
cos
/?,
cos
cos fi + z cos y (i) are the direction cosines of the
a+y
=/
perpendicular from the origin on the plane. And the equation of this perpendicular will be
cos
If
cos
//
cos
y
to the plane (r)
any plane
A^ + Br + Cs= D be parallel
we must
have
A
cos
B a
-
~~
C
cos
cos//
C
-
and
if
the line
v z
-~=
be parallel to the line
x
cos
at
cos ft
cos
-,
we must have
L
cos a
M
cos/3
=
N
cosy
Hence
the conditions that the line
M = -^
= D will
(54)
shall
be
perpendicular to the plane
ABC ==
A m
-
Ax + Ey + Cz
be
If the line
be in the form
v
)
we write
it-
xp
="
\~q
-
m
,
z =i
p
(55)
And
the conditions are
B
-
C
or
i
A
_,
;;/C)
\
)
= C
will
The equation
of a line passing through the point
.r
and
perpendicular to the plane
Ax + fy + Cz
B
=D
then be
xx yv zz
A

or in the unsymmetrical form
37
*-* = ~ (z-z)
yy
= ~(z-z).
$x+2j
\z
Ex. Find the equation of a line passing through the point

3) and perpendicular to the plane
63.
(i,
2,
5.
To find
the condition
that a straight line shall be parallel
to
given plane.
x
the form
Let the plane be b c a z y
Ax + l$y + Cz
=D
and the
line of
^- ~
AL +
Now
the
if this line is parallel to
normal
to the plane.
Hence
the plane it will be perpendicular to the required condition will be

.
BM+CN =
a straight
(56)
64.
To find
the conditions that
line shall coincide
plane
Ax + By + Cz
with a given
D.
x
L,
_y
TT~
_z c ~N~
:
The
line
must
fulfil
the
AL
-f
BM
-f
CN =
condition (56)
also
of parallelism above,
line as (a,
b,
o.
And
Bl>
any point on the
c)
Hence we must have the satisfy the equation of the plane. additional condition Aa + Cc o. + (57)
must
D=
2. Let
the equations of the line be of the form
x=mz +p y=nz + q
stituting these values of
,r
and
in the equation
of the plane,
Sub we
have
A(ms
whence
z
+/>)
+ K(nz -f q) + Cz =
t
D,
Am + B/* + C
52
JT
And
for coincidence this value of
must be indeterminate, and therefore A/
+ B^ D = Am + B + C =
(58) are the
conditions of coincidence.
NOTE.
This
last
method
is
ditions of coincidence of a straight line
a general one of determining the con and any surface given by its
That is substitute x and^ of the line in the equation of equation. the surface and since the z in the resulting equation mu^t be inde4
38
terminate
tity
if
NOTES ON SOLID GEOMETRY,

there be coincidence
we
treat this
and make the

To find
,
coefficients of the different
equation as an iden powers of z separately

,
equal to zero.
65.
the expression for the length
of
the perpendicular
its
PD from
any point P(x
,z
on a straight
lin.>
AB given by
where
equation.
i. Let
line
be cos
-=
cos
nates of any point
on the
line.
y Now PD
cos
a, b, c
2
are the coordi2

.
But
tion
PA of PA
(x
- a y + (v -by + (z -cY
a) cos
= PA AD and AD being
6)
(Fig. 13.)
the projec
on AB, we have
(x
AD
Hence
a+(yb)
cos /S+(z
cos y.
+ (z -b)cosy)\
2.
If the given line
(59)
be of the form
x
A
Then
cos
__
y
B
_zc
:
C
etc.
____
etc.,
And
therefore
PD
3.
If the given line
be
mz + p
nz + q
Then
PD
y=
2
66.
lines
To find the expressionfor the shortest distance between two straight given by their equations,
is
This shortest distance

the given lines
a straight line
(Fig. 14.)
AB
perpendicular to both
PB and
SR.
Let the given equations
x
co 5
c z = vb = ---
and -
cos//
cos;/
cos
xa yb -a cos -z cos
z
c
7
-,=
and 6
the
ft
angle between the
lines.
And
L,
M,
the direction cosines of the perpendicular
AB.

Then we must have
39
L
L
Whence
cos cos
a + M cos ft -f N cos y a + M cos ft + N cos ;/
o.
L
cos
/?
M
cos
cos
-cos /
cos
<*
CO3
y
cos
cos
cos
N
cos
<*
ft
cos
cos
ft
M
[(COS /S
2
-|-
COSy - COS /3 Cos y)*-t
^COS a COS
>
- CO*
a COS
2
>)
-j-
(COS a COS
^ -COBa COS
0)
I
2
1
sn
(Art. 15).
Now
(a
,
let
P be
the point
(a, b, c)
on the
line
PB and Q
be the point
AB
b\ c) on the line SR, itself, we have
Then
as the projection of
PQ
on
AB
is
AB = (a-a
/
)L + (6-6
)4-(6
)M + (c-c )N
=
(62)
(a- a )(cos^cosv _co!ycos/3
6 ) (cosa cosy - cosacosy )+(c -c ) (cosacos/3 -cosa cos^S)
If the given lines are expressed in other forms
cos
ft,
etc,
we can find cos a, from the given equations and substitute them in (62).
CHAPTER
V.
TRANSFORMATION OF COORDINATES.
67.
nates
To transform
to
= = r, NP = z be the coordinates of P referred to O and the axes O,r, Qv and Oz. Also let O be the new = a, A N = b, N O = c be its coordinates and let origin, and OA O H = x HK =_/ and KP = z be the coordinates of P referred to O as origin and axes parallel to the original axes. (Fig. 15.) Then .r == OA = OA + A A
the origin
,
of which referred Let OA .r, AN
to the
parallel axes through a old axes are a, b, c.
new
origin the coordi
or
Similarly
x
y
=a+ = b -f
c
x
)
(63)
and
Substituting these values in the equation of a surface we obtain the equation referred to the new origin and axes.
68.
To pass from a rectangular system

tlie
to
another system
tlie
origin
remaining
same.
;
,
Let Ox, Or, Oz be the old axes at right angles to each other O_v O/, Oz the new axes inclined to each other at any angle. (Fig. 16.)
OM = x, MN =.y, NP = z M N =.r NT = OM
=.<
,
Now
axis
Ox
the projection of the broken line + P -f is of the radius vector equal to the projection
a"
Let cos a, cos # cos Ox. new axes make with the axis O.*
,
on the on OM be the cosines of the angles which the
OM
MN N
OP
then
cos
=x
cos
a +y
a +
z cos
40

If cos
ft,
41
cos
ft
cos
ft"
be the cosines of angles which the new

y",
axes
the cosines with the axis Qy, and cos y, cos /, cos with Oz, we shall have similar values of the angles which they make and z. Hence the three equations of transformation are
make
x
j/
= =
=
x ^ x
cos cos
a +y
ft
cos cos cos
a +
/?
z cos
a"
cos
+y y +y
2
+ +
cos
z
>
/S"
(64)
cos
)
7".
We
have of course
cos
2
a + cos
ft
2
+cos 2 y
i
i
cos
a + cos
ft
+ cos* y =
>
(B)
cos
.r
For the angles A, /*, v between the new axes ofy and of ..v and y respectively we have
cos A
cos
fA
of 3 and
=
= =
cos cos
cos
r"cos
cos r
69.
cos
cos
+ cos /J cos or + cos // cos a + cos y5 cos

1
a"
ft"
+ cos y
cos ;/
J
/? ft
+cos y"cos y +cos ^ cos y
I
.
(C)
To pass from one system of rectangular coordinates

formulae in this case are the
to another also
rectangular.
The
same
as those in the last with the
exception that since the new axes are also rectangular cos A and formulae (C) give o, cos v
= o,
)
cos
//
=o
a
cos
cos
a"
+ cos
cos
ft
cos
ft"
+ cos y
cos
y"=
cos cos
a"
cos
cos
a +
ft"cos
<*
cos
ft
+cos ^"cos y cos /T +cos y cos y

ft
(D)
o.)
ditions,
Since between the nine quantities there are six equations of con (B) and (D) there are only three of the quantities, cos a,
ft,
cos
etc.,
independent.
70. In changing from rectangular axes to rectangular, there is another set of equations of condition among the quantities, cos a,
cos
ft,
etc.,
that the
new
equivalent to the preceding which result from the fact For a, a axes are rectangular. being the angles
,
a"
made by
the old axis of
.v
with the
2 2
new
2
rectangular axes,
etc.,
we
must have
cos 2 cos
2
cos 2
+ cos /? + cos ;/ + cos

tf
+ cos yS + cos 7/ + COS

<*
a"
2
ft"
2
y"
= = =
i
i
(E)
42
cos
cos
07v<:s
av SOLID GEOMETRY.
rr
<?
a a
/?
cos
/3
+ cos
cos
ft
cos
cos
cos
7 + cos 7 + cos
ft
cos ;/ cos 7
+ cos + cos + cos
<*"
cos
cos
y#"
"
7"
=o =o
o
>-
(F)
;
ft"
cos
7"
and the new coordinates expressed

.v
in
terms of the old are

\
.v
y
z
cos a +jcos ft +scos y ^cos a +ycosft +s cos y

"
(F)
)
x cos
+y
cos/?"
+s cos
7"
71. In the study of surfaces by sections made by planes it is often necessary to transform the coordinates in space to coordinates in To do this we must fix the plane with reference the cutting plane.
to the old coordinate planes.
as
z= A^ + By.
is
Then
the angle
Let the equation of the plane be given 6 which this makes with the plane
xy
determined bv the equation cos 6

cp
= =
and the
the axis of
o.
angle
which
it
traces
on
that plane
makes with
by the equation tan

Let
cp
^
,
the trace being Ajv +
Bj>
x Oy
be the given
Ox
to
it
which take
for the axis of
plane, cutting the plane xy in the line x and let a line perpendicular
CV
in the given plane be the axis
of v and
the coordinates of any point (V also let .v, QP

;
M in
=
v, r,
OR = .r RM =y
,
OQ =
the plane referred to the axes Ox\ P.AI =a z be the coordinates of
M
9
referred
to
the old
axes
Or, Os.
Then
the angle
MRP =
and
xOx = (p. Then PR =y
(Fig. 17.)
cos
(p
ft
PM =j/
y +y
sin
<p-RP
co^
(p.
.-.
z-y =x
.r
sin
cos
sin
cos 8 sin
gj V
)
y=
And
F(.v,
r,
(65)
<py
cos 8 cos cp
if
these values be substituted in the equation of any surface o the result will be a relation between x andy, coor z)
dinates of the curve cut from the surface by the plane.

72. If the cutting plane contain
one of the coordinate
axes, the
formulae are simplified and in many cases sufficiently general. Let x Oy (Fig. 18) be the cutting plane containing the axis of y .v Ox its trace in the plane ; zx the axis of x _r
;
PM =
OM =
.VOTES
ON SOLID GEOMETRY.
;
43
QP =
=. x, the coordinates of any point P in the section =.r, Then angle 2, the coordinates of P referred to the old axes. The formulae x cos , 0, and PQ x sin 0.
ON
.
NQ
PMQ =
MQ =
of transformation are
x
z
=x
x
cos
6*
^=/
sin
}
(66)
That
is,
we have only
to
make x
cos
0,
sin 0,
-=.y
in
the equation of any surface, in order to find the equation of the sec tion of this surface by a plane containing the axis ofj and making an
1
angle
with the plane xy.
CHAPTER
VL
THE SPHERE.
73.
To find the equation of the sphere. i. In rectangular coordinates.

,
Let
c
<,
be coordinates of the Centre, and Radius

is
1
=
c)
R.
z
The equation
or
if
then (Art. 1 )(x the origin be at the centre

2
aY + (y-l>Y + (z
2
.
(67)
^+/ + 2 = R
Let A,
/<,
(68)
2. Tn oblique coordinates. v be the angles of the axes then
the equation
is
(Art. 16)
x-a)(y-b)
cos A-f cos r
2(xa)(zc)
or
if
cos
)A+2(yb)(zc)
=R
2
.
(69)
the origin be at the centre
x*
.)
2
-s
+ 2xy cos
A -f 2AT2 cos
yu
+ 2yz
cos r
=R
(70)
3. In
Let r
tion
is
,
polar coordinates a, ft be the polar coordinates of the centre then the equa
2rr (cos ^ cos
rt
+ sin
sin
cos
(q>
>5))
=R
tf
(71)
0,
If the pole be at the origin
and the centre on the axis of
the
equation
is
r=2Rcos#.
Since that
positions.
is
(72)
the equation of the generating circle in any
one of
44
its

74.
45
The equation
JV
a
The Sphere under conditions (coordinates rectangular). (67) may be written

+>*
+
,s
2CLX
2by
21-2
+ 0* + &* +
or
.T
-f
/+
+ DA- + Ey + FZ + G
o.
(73)
And
sphere
ble)
since this
may be made
equation contains four arbitrary constants, the to fulfil four conditions (which are compati
and no more. Four given conditions give four equations for determining the constants D, E, F, G, and with these determined we know the radius and centre of the sphere, for we have only by
completing the squares to throw the equation (73) into the form
/ DV / ^\ / F\ D E F\x+-\ + \ y + - +(* + -) = + + -- C* 27 V a/ \ 2/
2
444
/
to see that the centre
is
(
D E F\ -- -- -,
, )
and the radius
is
2j
^
F~~
7"
1. The
v2
equation
2
of a sphere passing through a givtn point

2
d, e,
o.
f,
is
4-y +
+ D(^-^) + E(.y-^) + F(0-/)-^-^-/ =

is
(74)
If the given point be the origin the equation
=
2. The
equation
o.
(75)
of a sphere
cutting the axis of z at distances c
and
from
the origin is
x*+^+(z-c)(z-c ) + I)x + Ey= o

must give two values
condition.
for
z,
(76)
for"
^
fulfils
and
and
this
equation
that
3. The
equation
of a sphere touching
the axis
of z
at a distance
cfrom
the origin is
=o
>
(77) for this gives two coinci
dent values of z
"
when
J
=o
46
4. 7 he
origin.

equation
of a sphere touching
all three
axes at distance
a,
from
as
To meet
these conditions the equation
must be of such a form
to give equal roots for z

*
when
je=o
the
same equal
roots for
when
>
z=o
and the same equal roots
for
x when
a,
z=o
I
j
Let
the distance of points of contact from origin be

will
then the equation
be
=
as this fulfils the above conditions.
(78)
5. The
equation
centre on the axis
of a sphere passing through of x is
the origin
and having
its
* s +y + **=
6. The
is
2
2R.V.
(79)
equation
of a
sphere tangent to the plane xy at the point (a, b)
(;t--rt) -f
-)
+ * + F*=o
2
t
(80)
the plane xy.
for
then
z=o
gives
x=a, and y=b

of the expression
a point
in (a, b)
75.
Interpretation
(x-a? + (.y-b)* + (z-c)*-W.

i. Let
(x, y, z)
(i)
be the coordinates of a point
P without
the
sphere whose centre
is
tangent to this sphere at
and radius = R and let Then PM 2 = OP 2 the point M.

(a, b, c]
PM
be
.
OM
Now
and hence
PM =
2
(x-a)*+ (jy-) +
2
2
(z-<)
-R
2
.
Therefore the expression (i)

point
is
the square of the tangent from the
to the sphere.
(x, y z)
}
2. Let P
erect a perpendicular
PM
be a point within the sphere. to OP meeting the sphere R*
Join
in
OP
and
join
M, and
OM.
Then
PM =OM*-OP =
8
-((x-ay + (y-by + (z-c)*}

That
is the expression (i) becomes negative and represents the of the half chord through P perpendicular to the radius square
through P.
76. .Radical plane
of two
spheres.
is
Def.
The
radical plane of two spheres

to the
the plane the tangents
drawn from any point of which

If the equations of the
two spheres are equal.
two spheres are
the equation of their radical plane
is
For
point
this expresses (Art.

(_r,
75) that the squares of the tangents
from
y, z) to the two spheres are equal, and moreover it is an equation of the first degree in x, y and z and therefore the equation of a plane. If the spheres intersect their radical plane is their
plane
tres.
of intersection.
is
plane of two spheres
It may be easily proved that the radical perpendicular to the line joining their cen
77. ,The
point.
six
radical planes
of four
S
"
spheres intersect in a
common
Let S
spheres.
= o,
=o
S"
=o
o be the equations of the four
Then
the equations of their radical planes are
S-S =o
-S"
=o
=
o
= S-S = o
S-S"
"
"
S"-S"f:==o
These may be arranged in groups of four equations, which added vanish simultaneously and therefore the planes intersect in a common This point of intersection of the six radical planes is called point. the radical centre of the four spheres.
78.
Examples i. Find the centres and

:
radii respectively of the spheres
48

40
o.
\x +
5 jy
= o.
2. F ind the equation of a sphere passing through the origin and

the points
i, 2,
3,
i,
4,
5,
3,
i>
CHAPTER
VII.
CYLINDERS, CONES, AND SURFACES OF REVOLUTION.

79. CYLINDERS. Def. A cylinder is a surface generated by the motion of a straight line which always intersects a given plane curve, and is
The moving straight line always parallel to a fixed straight line. called \\\Qgeneratrix; the plane curve which it always intersects called the directrix or guiding curve.
80.
is
is
To find the general equation of a cylinder (coordinates rectangular}.

n,
i
Let m,
be the direction cosines of the

>
axis.
And
which
let
y
m
the axis.
its
nz
(i) be the equations of the generatrix in
and n are constant since the generatrix remains parallel to For convenience take the guiding curve in the plane xy,
equations will then be
~
.
\
5
(2)
Now making
=o
in
the equations (i)
we obtain
q)
x=p y
(3)
generator pierces the guiding curve
(x,
which the q for the point in y) in the plane xy.
Hence we have
and q between
(i)
F(/>,
= o,
and eliminating the arbitrages/
and
(3)
we obtain
?(x-mz,y-nz)
=o
(82)
the general equation of cylinders. If the cylinder be a right cylinder with its guiding curve in the in equation (82) m o, plane xy and the axis of z for its axis, then and the required equation of the cylinder is and o,
F(*, y)
= o.
(83)

8
1.
49
ders
We shall confine ourselves to cylin Cylinders of second order. whose equations are of the second degree. i. To find the equation of the oblique cylinder with circular base. R 2 = o. Hence (xmz, ynz) = o Here F(x,y) = x 2 +
gives (x
mzY + (y
nzf
z,
R =
2
o (84) the required equation.

If
is
2. To find
the equation
of
the right cylinder with circular base.
the axis be the axis of
the equation
=o J?(xy)
that
is
3. To find
the oblique cylinder with elliptical base.
Let the guiding
curve in plane xy be
a2
+^ bv*
=
J
i,
= o.
is
Then
(Jft jf)
x*
^
an ^ the equation
4. The equation
axis
is
the axis of z
of the right cylinder with elliptical base whose x* v*

is
F(x, y)
o,
that
is,
+^=
i.
5. The equation
curve
is
f = ^dx\
of the right parabolic cylinder whose guiding

o, is
$dx.
82. CONES.
Def. A cone is a surface generated by a straight line which passes through a fixed point and always intersects a given plane curve.
The
fixed point
is
called the vertex, the
moving
line
the generator,
and the given plane curve the

83.
directrix or guiding curve.
To find
the
general equation of a cone.

(a,
b,
Let the coordinates of the vertex be
c)
the equation of the

dircctrix
in
x
generator
^
its
~~n~
~T~
and take the ^1 r )

-
the
~
L
plane (xy)
eliminate
tions (i)
equation being then
(2).
Now
if
we
w and by means of the definition of cone and the equa and (2), the resulting equation will be the equation to the
z
cone, the locus of the right line (i).
Making
which
=o
5
in (i) the values
of* and
y,
namely,
~ a ~ mc = b nc 1 y
x
)
result will
be the coordinates of the point in which the gene-
50

xy and these
will
rator meets the plane
=o
the equation of the directrix.
We
consequently satisfy have therefore

(i)
(x,j )
F(a
me,
bnc) =
(3)
(3).
But from
m =^~
2
C
and therefore
becomes
zc
az
z
cy
ex
bz
or
zc
If vertex be
the general equation of cones.
on axis of
,
z,
then a
o
(86)
and
= o and
equation (85) becomes F(
\z
o.
84. Cone with vertex at origin.

If the vertex of the
cone
is
at
the origin
and the
directrix
it
in a
plane parallel to the plane xy,
and
be
~
at a distance c
from
then the
equation of the generatrix
will
= = =
I
.
(i) the vertex (o, o, o)
and the
directrix will be
v*^J
z
(2)
directrix
To
find the point in
which the generator meets the
we
i).
We
thus get
Hence we have
Therefore
(mc, nc)
= o,
r
but
m=
and n
from
(i).
<?,7
is
)=
(B7)
the equation required.
The equation
85.
(87)
is
homogeneous equation
1. The
equation
in
x,y and
z.
Cones of second degree.
of an
2
oblique cone with
circular base.
The equation
Hence
of the directrix
is
~F(x, y)
= x +y R =
2
o.
zc
zc
zc

or
(az
5!
-ex)* + (bzcy}^
K\z-c}\
(88)
2. To
z being the axis
find the equation of a right cone with circular base, the axis of of the cone and vertex being (o, o, c). The equation
.
of the directrix
F(*,.> v
is
)=-*
+y-R =o
s
c.
Hence
W ^ )=o -^ 7^-R = zc
is
<7
or
2
jt-
+y=
(2
r)
(89).
This
is
a cone of revolution about the
axis of z.
3. The
elliptical,
equation of a right cone with vertex at the origin or hyperbolic bases.
and circular,
The equations
of the circular base (directrix) are
Hence
/ ex
\T"
cy\
fix
*~z*
c^y
""
2 ^
~z)~
2
""?"
The equations
of the elliptical
and hyperbolic
directrices are
^ +Hence
_ =o z=c]
i
)
and
| --|
2
- = z=
1
respective]y
c)
the cones are
x f z =oor--+--=-5
9
2
. .
(91)
c*x*
86.
SURFACES OF REVOLUTION.
To find the general equation of a surface generated by the revolution of a plane curve generator about the axis of z.
Let SPi=r be an ordinate of the point
to
the axis of z of the
52
plane curve and
Then SP 2
That
= ON
the
OM = x, MN OM + MN
2
=y,
,
NP = z
.r
2
the coordinates of P.
2 or r
+y
(gen
is,
distance from any point of revolving curve

is
erator) from the axis of z
r==vC?4j? (i). But r being an ordinate of the generating curve to the axis of z we must have by the
equation of the curve in any position r

eliminating the arbitrary r between (i)
(z)
(2).
Therefore
and
(93)
(2)
we have
V^+y = F(s)
the required equation of surfaces of revolution about axis of z. If the curve revolved about the axis of .r the equation is
Vy +s
87.
F(JC).
(94)
Surfaces of revolution of second order.

revolution
i. Equation of Cylinder of
equation of the revolving line
is
= a.
gives
about the axis of
z.
The
V.
-**+
x-+f
a\
at
2. Equation of a Cone of revolution about the axis of z, vertex The equation of the generating line is r = m(zc}. (o, o, c).
Hence
.#
+y
the tangent of the angle
m*(zc)* (95) the made by side
required equation where of cone with axis of 2.
is
3. Equation of the Sphere. 2 a* or r =\/a* z
The equation
.
of the generating curve
Hence
4.
about
Equation of the Surface generated by the revolution of an

its
ellipse
conjugate axis.
r*
is
The
generator
- +Z -=
or
r~
= ~(b
is
/70
OX 2
).
Hence
the equation of the surface
+=This
is
(96)
one of the
ellipsoids
of revolution called the
oblate spheroid.

5
.
53
ellipse
Equation of the Ellipsoid generated by the revolution of an

its
about
transverse axis the {Prolate spheroid].
Take
the axis of
is
x
jv
as the axis of revolution.
Then
the equation
of the generator
+ 75
or r
=
a
>(#
2
x*)>
Hence
vj + s*
*
_
x
F(,r) gives
v
i
r**;
it
(97)
the required equation.
revolves about
lution
88. Hyperboloids of revolution. its conjugate axis
Definitions.
When
the Hyperbola
generates the Hyperboloid of revo

it
of one
sheet.
When
erates the hyperboloid
it revolves about the transverse axis of revolution of two sheets.
gen
i. Equation of the Hyperboloid of one

r
2
sheet.
Let the axis of z be

8 2 ).
the conjugate axis then
-273
or r 2
= j3(s +
a2
Hence
2. 7%f
the axis of
equation
of the Hyperboloid of revolution of two
sheets.
Take
as the axis of revolution.

a:
2
Then
^ 2 ).
the equation of the
r2
generator
is
2 or r
b*
(j\r
Hence
for the
equation of the surface we have
A-
,-
?--?-=
89.
(99)
Equation of the Paraboloid of revolution about the axis of\.

2
The equation of the generator is r 2 \dx. Hence the equation of the Surface isjy 2 + 2
5*
\dx.
(100)
CHAPTER
VIII.
ELLIPSOIDS, HYPERBOLOIDS,
To find
the equation to the surface
is
AND PARABOLOIDS.
Ellipsoid.
89.
of an
generated by a variable ellipse which always moves parallel to a fixed plane and changes so that its vertices lie on two fixed ellipses whose planes are perpendicular to each other and to
Def. This surface
the plane of the
moving
zx
;
ellipse,
and which have one
axis in
common.
Let BC,
of
CA
;
>
(Fig. 19)
0,
be quadrants of the given fixed

c their
ellipses traced
in the planes
OC
common
semi-axis along the axis
b (on the axis of y) the a (on the axis of x), and OB z, a quadrant of the variable generating ellipse other semi-axes QPR in any position, having i s centre in OC and two of its vertices in the
;
OA =
ellipses
AC, BC, so
that the ordinates
QN,
RN
are
its
semi-axes
also let
ON =
:
z,
NM = x, MP =y be
y*
the coordinates of any point
in
it
Then
3? -^
+ .i
=
-.
i.
And
since
Q
i
is
on the
ellipse
AC we
have
a
Similarly
--
u
2
Hence eliminating
RN
and
QN
we have
*
-r +the equation to the surface.
90.
^L
7,-+--r==
ellipsoid
,
(loi)
To determine
.v
2
-f-
the form
2
of the
from
its
equation.
Since in
y
y- -f
z*
the equation
can only receive values between a

54

and
is
55
c,
a,
y between b and
all directions.
b,
and
V
-f
between
and
the surface
limited in
If
we put
on
xy,
o we obtain
is
A2
r=
i,
for
the equation
to
the
trace
If
which
therefore the ellipse
AB.
AC.
BC.
we put y we put x
o we have ~--\
x*
z1
^-
=i,
the ellipse
If
= o we
have
-f
i,
or the ellipse
These three sections by the coordinate planes are called the principal sections, and their semi-axes a, 6, c, are the semi-axes of the ellip soid and their vertices the vertices of the ellipsoid, of which it has six.
;
If
we make z=h we have
the equations of any section parallel to xv, which is an ellipse similar to AB, since its axes are in the ratio of a to b, whatever be the value
of
h,
ner
to
In the same man c. and which becomes imaginary when h sections parallel to xz and yz are ellipses respectively similar AC and BC. The whole surface consists of eight portions pre
>
all
cisely similar
and equal
to that represented in the figure.
Cor. If
=0
yA
the ellipsoid
becomes
z,
^_
j,2
z
t--
the ellip
soid of revolution about the axis of
Art. (87), all the sections of
which by planes parallel toyz, are circles. Hence the spheroids may be generated by a variable circle moving as the variable ellipse, in
Def. Art. (89).
91.
To find the equation
to the is
hyperboloid
of one
sheet.
generated by a variable ellipse, which moves parallel to a fixed plane, and changes so that its vertices rest on two fixed hyperbolas, whose planes are perpendicular to each
Definition.
This surface
other,
and
to
the
having a
Let
common
and
plane of the moving ellipse, the two hyperbolas conjugate axis coincident with the intersection of
their planes.
(Fig. 20.)
OC =
of z
;
AQ
c their
OA = a, OB =
the given hyperbolas traced in the planes zx,yz ; common semi-conjugate axis coinciding with the axis
b the semi-transverse
its
BR be
axes
QPR
the generating
ellipse in
any position having
plane parallel to xy, its centre in
56

1
OC, and
its
vertices in the hyperbolas

its
AQ, BR,
so that the ordinates

2,
NQ, NR,
ellipse
are
semi-axes.
Also,
let
MN = x, MP = y, ON
;
be the coordinates of any point
in the generating ellipse
then the
PQR
gives
2
"NQ
+ NlT =
2 2
Also from hyperbola
AQ
BR
NQ
2
z*
"~=i.
And from
Hence,
hyperbola
--
NR
2
z"
i.
f
8 A-
or
V
^-
2 -
(102) the equation to the surface.
92.
To determine
the
form of
the hyperboloid
of one
sheet
from
its
equation.
Since the equation (102) admits values of x,

negative however
sides of the origin.
~z-
and
z positive
and
all
large, the surface
is
If
we put
extended indefinitely on
o we obtain
is
+ ~i
for the trace
on xy which
the ellipse
AB.
r
2
Similarly
s
2
the sections by the planes
xz and j 2
are respectively
^
The
the
hyperbola
AQ, and
v
1
-^
the hyperbola
BR.
ellipse
AB
and the hyperbolas AQ and BR are the principal sections. The sections The parallel to xy are all ellipses similar to and greater than AB. sections parallel to xz and yz are hyperbolas similar to the principal
sections.
The
and
c
semi-axes a and b are called the real semi-axes of the surface

the imaginary semi-axis,
i.
since
x
is
=
real
o and
axes
=o
give z
=
is
c\/
The
extremities
of
the
are
called
the
vertices of the surface.
The
one
surface
continuous and hence

space in
called the hyperboloid of

rior of the
sheet.
The hollow
the inte
volume of
this
hyperboloid of which the ellipse
AB
is
the
smallest section has the shape of an elliptical dice-box.

Cor.
If b
57
a the equation becomes

fl
=
2
j
c
that of the
to
hyperboloid of revolution of one sheet.

are all circles.
93.
Its sections parallel
xy
To find
the equation to the hyperboloid
of two
sheets.
Definition.
This surface
itself,
is its
moves
tices in
axis.
parallel to
with
generated by a variable ellipse which axes on two fixed planes at right an
gles to each other
and
to the plane of the generating ellipse
and ver
two hyperbolas in those planes having a
common
transverse
xy,
AQ and AR be the given hyperbolas traced in the planes zx, OA = a their common semi-transverse axis along the axis of x, OB = b OC = c the semi-conjugate axes along the axes ofy and z QPR the generating ellipse in any position having its plane parallel loyz its centre in Ox, and its vertices AQ, AR so that the ordinates MP = z be the QN, RN are its semi-axes. Let ON = x, MN
Let
; t
=j>,
coordinates of any point
in the ellipse.
(Fig. 21.)
"
also from hyperbola
AQ
AR
and from hyperbola
"-
l>
a1
i.
Hence
>-?
the equation to the surface.

94.
tion.
To determine
the form
of the hyperboloid of two
sheets from
its
equa
The equation shows that all values of x between -f# and a give imaginary results, therefore no part of the surface can be situated be tween two planes parallel toyz through A and A the vertices of the common transverse axis ; but the equation can be satisfied by values
58
of x,
to
>

,
z,
indefinitely great, therefore there
is
no
limit to the distance
which the surface may extend on both sides of the centre.

If
V* we make x-= o we have ~-
Z*
-\
=
For
o
for the principal
section by the plane yz.

z*
-
h and h
ellipses.
>
we have
v* -\
2
//
^-
which represents similar
The
tively.
and AQ, respec principal sections by the planes xy and zx are For the sections parallel to xy and putting z I we
AR
have
in
= +
i
a hyperbola similar to
AR
with
its
vertices
the opposite branch of that hyperbola and conjugate axis In the same way the sections parallel to zx are hy similar to with vertices in and its opposite branch perbolas and conjugate axes parallel to Qz, 2a is the real axis of the surface
parallel to Oy.
AQ and
AQ
AR
and
its
vertices
the vertices of the surface.
The
axes 2b and 2c
are the imaginary axes of the surface as it cuts neither y nor z. The whole surface consists of two indefinitely extended sheets perfectly
similar
and equal, separated by an
interval.
Hence
its
name.
the equa
Cor.
\ib-c
the equation
becomes
~
^
tion to the hyperboloid of revolution about
its
transverse axis.
95. Asymptotic cones
to the
two hyperboloids.
sheet has
r
2
i. The
Putting
hyperboloid
of one
x*
an
z*
-j-
interior asymptotic cone.
its
equation
T +
^
.
(i)
in the
form
4.
-=s
/I ^---
(2)
Now when
is
very great
,-
is
very small, and hence the limiting form of (2) for z increased
is
without limit
r
its
v + ~- =
(3)
the equation of an elliptical cone

its elliptical
having
vertex at the origin
and
section parallel to xy.
Moreover,
this elliptical section is
always within the corresponding

section of the surface by the
(i)
59
z
same
plane.
For putting
h in
and (2) respectively we have
~-?1
+
+
~rr
//
= +
l
c*
f r
tne section of the surface
x
a
^2
4^1
b
/ =
/r
72
c
for the section
of the cone.
This cone
2.
asymptotic to the hyperbola. The hyperboloid of two sheets has an exterior asymptotic
is
cone,
x~
Putting the equation
a2
V* ^-r
(j
z~
r r
(i)
under the form
we have
as a limiting
form of
this
jv
equation when jy and z increase without limit,

2
2
j -----|
---
z~
(2)
an
elliptical
cone with vertex
at the
origin
and with an
this elliptical
elliptical
section
is
section parallel to the plane \z. Moreover, greater than the corresponding section of
plane.
2
2
the surface by the

respectively
same
i
For putting
--
.h in (i)
and
(2)
we have
-,*
H---
=
55
h*
5 2 i
^
This cone
96.
is
"T"
?"
asymptotic to both branches of the hyperboloid.

the equation to the elliptic paraboloid.
To find
Definition.
This surface
lies
is
generated by the motion of a parabola
whose vertex
two parabolas being perpendicular to each other, their axes parallel and their con cavities turned in the same direction.
fixed parabola, the planes of the
on a
Let
OR
axis along the axis of
be a parabola in the plane xy, its vertex at the origin, its RP the generating .r, and / its latus rectum
;
parabola in any position with
its
and axis
parallel to O.v,
r,
and
z
let
plane parallel to zx, vertex in OR, / denote its latus rectum. Also let
it
:
NM = MP also draw RM parallel to

ON
x,
be the coordinates of any point P in Oy. (Fig. 22.)
60

z*
Then
= I RM = / .M N and / = .OM
.
j-
+ -j- = x
IO 3) the equation to the surface.
97.
To determine
the form
of the
elliptic
paraboloidfrom
its
equation.
Since only positive values of x are admissible, no part of the sur face is situated to the left of the planers. But the surface extends 2 Ix indefinitely in the positive direction of x. If we makej o, 2
the equation to the principal section OQ, and to zx are parabolas equal to OQ, with vertices in
is
all
sections parallel
;
OR
similarly, all
sections parallel to xy are parabolas equal to the other principal sec tion OR, with vertices in OQ. If we make x h we have
21 + Jl-, ~
Ih
Ih
Therefore the sections parallel to zyare similar
ellipses,
and hence
its
name.
Cor.
If
/=
the equation
becomes
2
j
+ =
2
/.v,
the paraboloid of
revolution.
98.
To find
/he equation to the hyperbolic paraboloid.
Definition.
This surface
lies
is
generated by the motion of a parabola
whose vertex
on a
fixed parabola, the planes of the
two parabolas
being perpendicular to each other, their axes parallel, and their con cavities turned in opposite directions. (Fig. 23.) Let OR be a parabola in the plane of xy, vertex at the origin, and
axis along with the axis of x,
and
/ its latus
rectum,
RP
the generat
to O.v,
z,
and
ing parabola let t denote
in
any position, vertex in

its
OR,
axis parallel
x,
latus rectum,
and
it
;
ON =
draw
the coordinates of any point
in
NM =y, MP RM parallel to
;
Oy.
Then
z-=!
but
MR and/=/.OM OM - MR = ON =
.
AT.
Hence
99.
V
:
2 ---- = x
(104), the equation of the surface.
To determine
the form
of the
hyperbolic paraboloidfrom
its
equation.
The surface cuts the coordinate axes only at the origin, and since the equation admits positive and negative values of.r, v, s, as great

as
6f
the
we
If
please, the
surface extends indefinitely both ways from
we makejy = o we have z~ = I x the principal section, the para OQ, with its concavity turned towards the left ofyz, and all sec tions parallel to zx are parabolas equal to OQ with their vertices in OR. Making z = o we have y 2 = Ix the parabola OR, and sections
bola
parallel to
If
origin.
xy
are parabolas equal to
OR
the
with vertices in
principal
OQ.
in
we
make x
we
have
section
yz,
z^ I = y +Jl
tions parallel
-
or two straight lines through the origin; and for sec
toyz making
h we have
its
-2
=
2
-
a hyperbola with
vertices
in
OR, and con
jugate axis parallel to Oz.
For h negative the section becomes

its
In
7-
Ik
a hyperbola with
vertices in
OQ, and conjugate
axis parallel to O.v.
The
surface has but one vertex,
and consists of one sheet and one
infinite axis.
100.
Asymptotic planes
v*
to the hyperbolic
paraboloid.
The equation
z
--==
z*
-jr
=
/
T
x may be
l
written
2
j
z*
-j~ /
~jr /
+ ~7 z J
x\
)
which has
for its limiting
form
Z
<1
when
or
-:
y
=
and
become
infinitely great with
regard to x,
v
=.
7-
-.
z
.
This represents two planes
= H--z
and
= --^ V?
through the origin and asymptotic to the surface.

to all the hyperbolic sections of
These planes contain the asymptotes

the surface parallel toj 2.
101. The
ellipsoid
elliptic
and
to
hyperbolic paraboloids are particular cases
and
hyperboloid
of one
sheet respectively
when
the centres
of the of these
surfaces are removed
infinite distance.
Take
the equation
+ ^-
=i, and
transfer the origin to
62
the
left
NOTES ON SOLID GEOMETKY.

vertex of the axis
20,
(a,
1
.
o, o).
(New
coordinates being
parallel to the primitive.)
-,.
or multiplying through by a
-
vJ1
or a
o 2
r2
--
y*
t-
=2jr
-y^-
(i),
in
which
a
in
and
a
zx.
are the semi-latera recta of the principal sec-
tions
xy and
Now make
/
a =. oo
/
and put
V
-
c*
and
a
which remain
finite,
equal to
and
respectively.
.*.
(i)
becomes
2x, the equations to the paraboloids.
=y1
JT
02.
been studying are
The equations of the surfaces of of the two forms
the second order
which we have
=D
=kA*
and we
(i) (2)
will show hereafter that all the surfaces of the second degree transformation of coordinates be included in these two forms. may by The first form (i) includes the sphere, ellipsoid, hyperboloids,
elliptical
.r,
-y,
cones of second order, have centres. For if
and hyperbolic cylinders

z be written for
which
(x, y, z) in (i)
the equation is not altered, therefore for every point P (x, y, z) on the surface there is a point P ( x, y, z) and PP passes through the origin and is bisected in O.
Moreover, the coordinate planes bisect
all
the chords parallel to
the axes perpendicular to these planes respectively and are principal planes of the surface.
The second form

loids
(2) includes the elliptic and hyperbolic parabo and the parabolic cylinder which have a centre at an infinite
distance.
The planes ^.s: and zx are principal planes of the two paraboloids, the other principal plane being at an infinite distance. Also both families may be represented by the equation
NOTES ON SOLID GEOME IRY.
63
the origin being at the vertex no centre.
and
A=o
when
the surfaces have
EXAMPLES.
1.
Construct the sphere whose polar equation

r
is
=a
sin
6 cos
cp.
2. Find the locus of the point the sum of the squares of the dis tances of which from n fixed points is constant. 3. Find the locus of the point the ratio of the distances of which
from two fixed points is constant. 4. Find the equation of the surface generated by the motion of a variable circle whose diameter is one of a system of parallel chords
of a given circle to which the plane of the variable circle
dicular.
5.
is
perpen
The sphere can be
represented by the simultaneous equations
= a cos cp cos y = a cos cp sin z = a sin cp

x
6.
6 6
>
The
ellipsoid
may be
represented by the equations

a cos
<p
cos 6
y
z
7.
= c sin
b cos cp sin
cp
The hyperboloid
of one sheet
may be
\
represented by the equa
tions
y
z
8.
=c
a sec cp cos 6 b sec cp sin 6

tan gj
The hyperboloid
of two sheets
may be
\
represented by the equa
tions
y
z
9.
a sec cp b sin 6 tan cp c cos 6 tan cp
>
line
moves so
its line.
that three fixed points
fixed planes mutually, at right angles.
on it move on three Find the locus of any other
point
P on
64
Solution
:
Let the three fixed planes be the coordinate planes (x,y,z) the coordinates of P. A, B, C the points in which the line meets the coordinate planes of yz,
xz,
xy, respectively.
=:<p,
Take
PA=<z,
PB=/>,
<ACA
<CB
.r=:0
(CA being
the projection of
sin 6,
PC=c, ON=.r, NQ=,y, QP = z, CA on the plane xy and B

sin
q>,
the projection of B on the axis of x). Then x=a cos cp cos Q,y=b cos
face
is
z= c
and therefore the sur
q>
an ellipsoid.
10.
is
equal to
11.
all
Find the locus of a point distance of which from the plane xy its distance from the axis of z (coordinates rectangular). Find the locus of the centres of plane sections of a sphere
pass through a point on the surface. Find the equation of the elliptical paraboloid as a surface
which
12.
generated by the motion of a variable ellipse the extremities of whose axes lie on two parabolas having a common vertex and common axis and whose planes are at right angles to each other.
13.
in a similar
14.
Find the equation of the hyperbolic paraboloid as generated manner by the motion of a variable hyperbola.
Construct the surface r sin 6
= a.
15.
Find the equation
to the surface B
= JTT
in rectangular coor
dinates.
CHAPTER
IX.
RIGHT LINE GENERATORS AND CIRCULAR SECTIONS.

103. SURFACES of the second degree admit of another division, viz. those which can be generated by the motion of a straight line
into those
into
and
This property which we have seen to and cone we shall now show to belong also to belong The the hyperboloid of one sheet and the hyperbolic paraboloid.
which cannot.
to the cylinder
;
ellipsoid being a closed finite surface does not possess this property nor the hyperboloid of two sheets, since that consists of two surfaces
separated by an interval limited in one direction.

104.
nor the
elliptical
paraboloid, since that
is
Straight
line
generators
of the hyperboloid of one
sheet.
The equation
a2
a
vz
2
of the hyperboloid of one sheet

z*
5i
+
fx
\a
^
y
o*
c
mav be
.r
2
-
z*
written
a2
c
-
-r-
=
(
~-~
r
1
z\ fx
) \a
z\
c
f
\
y\(
b
y\
b)
\\
) \
Now
(A)
is satisfied
by the pair of equations
+ 7
and also by the pair
=K
V\
(B)
i)J
66
And m
lines,
and
being arbitrary equations (B) represent a system of straight all of these lie on the hyperboloid as the two equations
together satisfy the equation to the hyperboloid.

Similarly equations (C) represent another and distinct system of straight lines which also lie on the hyperboloid which is the locus of
both systems, and we shall see the lines of either system

as generators of the surface.
105. No tivo generators of the same system For example take two of the system (B),
/-v
i
may be used
intersect one another,
A
c J
)
= i
v
-.
\a
x
a
(0
(2)
Combining
the
first
equation of (i) with the
first
of (2)
we obtain
(in
m")
=
J
Combining
have
the second equation of (i) with the second of (2)
we
or v
These values
1
for
being incompatible the lines do not

the system
intersect.
06.
Any generator of
(B) will
intersect
any generator of
the system (C).
Take
,
(*--*-}= c /
(x +
t-Z b
of system (B)
,
z \
r)
=w
+ y\
i)
67
(4)
of system (C).
Eliminating x, y,
and
a;
we
obtain the
identity
m"
m",
therefore the lines intersect.
Hence, through any point of an hyperboloid of one sheet two straight lines can be drawn lying wholly on the surface.
107.
No
if
straight line
lies
one of the systems
of generating
on an hyperboloid which does not belong lines (B) or (C).
to
suppose a straight line H to lie entirely on the must meet an infinite number of generating lines of Let two of these (one of B and one of both systems (B) and (C).
For,
possible,
it
hyperboloid,
C) intersect
in
two different points,
\ve
could then have a plane
in
tersecting the surface in three straight lines,
which
the equation is of the second degree. lie on the surface.

1
Hence
impossible since no such line as can

is
08. The hyperboloid of one sheet
may
be generated by the motion

lines
straight line resting
on three fixed straight

to the
it
which do not
of a intersect, and
which are not parallel In the

first
same plane.
place
is
which
tions.
is
to generate a
its
necessary that the motion of a right line surface should be regulated by three condi
For, since
equations contain four constants, four condi
position absolutely ; with one condition less the position of the line is so far limited that it will always be on a certain locus whose equation can be found.
tions
fix
its
would
Take then three fixed generating lines of the system (B), these do not intersect, nor are they parallel to the same plane. Now, if a straight line move in such a manner as always to intersect these three
straight lines,
it
will trace
out the hyperboloid of which they are the
generating lines.
For the moving line meets the hyperboloid in three points (one on each of the fixed straight lines), and hence must necessarily lie wholly upon the surface. For the equation of intersection of a line and this surface being a quadratic equation, if satisfied by more than
two
line,
roots,
it is
satisfied
its
by an
infinite
number.
The moving
straight
therefore, in
different positions, will generate the
hyper
boloid.
68
109.

Lines through the origin parallel and (C) lie on the cone
x*
respectively to
generators of the
systems (B)
y
-73-
z=
asymptotic
to the
hyperboloid.
For
this
equation of the cone
may be
put in the form
A A \a + A = _,vz, A
\
c J
c t
which gives two systems of lines through the origin lying on the cone, one system evidently parallel to the lines (B) and the other to the lines (C).
no. The
projection
is
of a generating
to the traces
principal planes.,
tangent
line of either system upon of the surface on those planes.

is
the
The equation
of the trace of the surface on the plane zx
The
projection of the line of system (B) on xz
f_ i\ + * +
\
=2m
c
or
^}. 2m
*
a
+ 1=*. i = 2m c
"V
(I)
P
~|
Now.
the condition that a line in the form
=
T
shall
be
p
tangent to the hyperbola
q
c*
~
.r
2
z1
F =I
a*
IS
~^~
This -condition
is fulfilled
by the projection
(i), for
+!)
(i-m
Y
tangent to the hyperbola.
Hence
III.
this projection is
line
The straight
generators of the hyperbolic paraboloid.
The equation
of the hyperbolic paraboloid

=y
69
--- = x may
be written
And hence
it is
satisfied
by the pair of equations
V7
n + ^)
or by the pair
=i
_
HE).
Hence the surface has two systems of and (E).
straight line generators
(D)
The lines of both systems are parallel to the asymptotic planes of the surface respectively. The equations of these planes being
z
==o
and
112.
We
and (35)
that
can show in the same manner as in the Articles (34) no two lines of the same system intersect and that a
;
line of either system intersects all the lines of the other system, and that no other line than the lines of these two systems can lie on the
hyperbolic paraboloid.
surface two lines
And
as in (108) of a straight line which rests on two fixed straight lines and is con also by a straight line which rests stantly parallel to a fixed plane
;
And hence that through every point of the may be drawn which lie wholly on the surface. that this paraboloid may be generated by the motion
on three
fixed straight lines
which are
all
parallel to the
same plane.
113. The projections of the generating lines on the principal planes are
tangent
to the
principal sections
The
principal section in
of the paraboloid. xy is y* = Ix (i).
NOJÊS
The
ON SOLID GEOMETRY
xy
is
projection of any line of the system (D) on
/t
/ /
j=
mx
m
-WJ
or y
O 2
jp-fJ^-rw
2m
( *
2/ )
Now
the tangent line to the parabola
y>=
Ix
is
of the form
y
Hence
114.
lines.
tx
+
4^
and
if /
=
2
then
4/
= -5
2m
-.
the projection (2)
is
tangent to the section
y = lx
Distinctions
of surfaces of second order generated by straight
All generators of the cone intersect in one point. of the cylinder are parallel. Hence cones and In the case of the cylinders are called developable ruled surfaces.
the
All
the
generators
hyperboloid of one sheet and the hyperbolic paraboloid, the gen

erators
of
neither
system
intersect
or are
parallel.
These are
styled skew ruled
surfaces.
The
distinction between these last two
surfaces
is
that the generators in the paraboloid are parallel to a fixed
plane.
115. Plane sections
If
2
aces of the second order. of surf
we
intersect
2
the surfaces represented by the general equation
A* + By + Os + 2h. xz +
2
2B>
+ 2Cxy + 2& x+2 B"y+2C"z
by the plane z
2 2
o we
will obtain
A^ + B| + 2C A7
If
|-2A"jt:
+ 2B
=D
(i) a conic section.
we
intersect
it
by a plane z
a we have for the curve of inter
section
Aa 2 -h B^ + 2C xy+ 2G x + 2H>=D
2
a conic similar to the conic (i).
Therefore sections of surfaces of the second order by parallel

planes are similar curves, and hence, in determining the form of these sections we may confine ourselves to the discussion of sections through
the origin.

1 1
^
of
6.
To determine
the nature
of
the curve formed by the intersection
a surface of the second order by any plane.
Take
the equation
= 2A
Make
,r.
curve of intersection in
its
And own
x
in order to get the equation of the
plane
cos
q>
-f
y cos
y
sin cp
y
z
zz:
x
f
sin cp sin 6.
cos 6 cos cp
=y
2
See Art. (71).
Arranging the
result
we have
cp)
*(K cos
2
2
r/>
+B
sin
+ 2x y
(K
2
B) cos 6 sin
cp
cos cp
,
the equation to a conic section which will be an ellipse, parabola or hyperbola, (including particular cases of these curves,) according
as the quantity
2 2 2 B) cos 6 cos
(A
cp sin
cp
2 (A cos
<p
+ B sin + B cos #cos

2
2
<p)(A
cos 6 sin 2 cp
<p
+ Csin
(i)
0)
or
is
-AB
Hence
cos
cp-AC
cos
sin
BC
sin
2
<p
sin
8,
negative, zero or positive.
every section of an ellipsoid
is
an
ellipse
because A,
B and
are all positive.
The sections of the hyperboloids may be ellipses, parabolas or hyperbolas since one or two of the quantities A, B and C will then
Hence for the elliptic paraboloid in For paraboloids A = o. B and C have the same signs the section is an ellipse except when B = o or cp = o in which cases it is a parabola. For the hyperbolic paraboloid since B and C are of contrary signs the section is a hyperbola except when 6= o or cp=o when it is a
which
;
be negative.
parabola.
1 1
7.
it
Circular
sections.
Since the section
is
referred to rectangular
axes
cannot be a
circle unless the coefficient of
xy
vanishes
72
or

(A
#
2
B) cos 6 sin
7T
cp
7T
cos cp
or cp
or
or
&=
o
be
dicular
1 1
which shows thaty^r # circular section the cutting plane must to one of the principal planes of the surface,
8.
perpen
Let us
now examine
the surfaces of the second order for cir
cular sections.
Take
first
the surfaces having a centre
and
therefore represented
by the equation
A* + B/ + Cz =
2
2
i.
(i)
Since every circular section must be perpendicular to a princi pal plane, let the cutting plane contain the axis of y, and make
the angle 6 with the plane
xy
To
transform (i) to this plane
make
cos 6
x
z
=x
x
2
y=y
sin 6.
Art. (72).
Hence we have
.v"(A
cos
+C
sin
0)
+ B/* =
(2)
which represents a
circle if
Acos #-fC
2
sin
B
.
or
tan
= B-A
(3)
it
We
must now examine
for
each of the surfaces which axis
is
that coincides with the axis
ofj>.
i. For
the ellipsoid
A=
3=-^-,
Hence
parallel
for a real
B b must
lie
(in value)
between a and
or the
is
axis of the surface to which the cutting plane of circular sections

is its
mean
axis.
2. For the hyperboloid of one
sheet since
we cannot have B ne-

gative
we must put
= ^
>
a or the cutting plane
is
parallel to the greater of the real
axes.
3. For
the hyperboloid of two sheets since
we cannot have
and
negative,
we must put
.*.
>
or the cutting plane
is
parallel to
the greater of the
im
aginary axes. Since tan 6 has two equal values the cutting plane may be inclined 6 to the plane of xy. at an angle 6 or 180 Hence there are two
sets
the surface
of parallel circular sections of the surfaces having becomes one of revolution we have tan 6
a.
centre.
oo or o,
If
and
and
the two positions of the circular sections coincide with each other, are parallel to the two equal axes.
119. Secondly.
For the surfaces not having a
centre,
we take
2 equation B_/ + Cs = 2k x (i). i. For the elliptic paraboloid, B and C have the same sign. 2 2 2 2K x cos 6\ and hence Transforming (i) we have B/ + Or sin # = B, or for circular sections we must have the condition C sin
2
sin 6
- Therefore the cutting plane

whose
latus
is
perpendicular to the
principal section
rectum
is least.
2. For
signs, sin
the hyperbolic paraboloid, since
6
in
is
tersect
it
B and C have different and no plane can be drawn which shall in imaginary, a circle. This was evident, too, from the fact (Art.
hyperbolic
116)
tions.
i
that
the
paraboloid
can
have
no
elliptic
sec
20.
Then,
7
to
sum
up,
all
the surfaces discussed with the excep
tion of the hyperbolic paraboloid
admit of two
sets of
planes of
cir-
74
cular sections.
variable circle

whose centre
Therefore they can be generated by the motion of a is on a diameter of the surface.
121. The planes of circular section the equations of the surfaces, as follows:
may be found
directly
from
The
equations of the central surfaces
may be
or
written
which shows that either of the planes

BJI
<\/A
+V^-^
=o
same
(i)
<v/A
B.r
it
VB
C.s
o (2)
cuts the surface in the
line in
which
cuts the sphere
Hence
the planes (i) and (2) and

circles.
all
planes parallel to them cut

in a similar
the surface in
The equation to the elliptic paraboloid may be treated manner, thus showing its planes of circular section.
122. Sections
of Cones and
Cylinders.
i. The
sections of the cones
may be
inferred from Art. 95.
For
elliptic cones, sections of the
hyperboloids by any plane are always
similar to the sections of the asymptotic cone to the surface the same plane, as is evident from the equations respectively.
made by Hence
the section of a cone of revolution by a plane will give an ellipse, But we will examine this case more par parabola, or hyperbola.
ticularly.
In the equation of the cone of revolution

2
A. +
/= L
J
_, ),
put
or
.v>+y
(*-<)
(when
x=x
tan v J
cos 0}

And we
jr
2
75
have for the curve of intersection by the plane containing
the axis ofjy

2
(cos
#tan 2 v
sin
6)
+ _/*
tan
v+2cx
sin
6c = o
1
(i).
This equation (i) represents an ellipse, parabola, or hyperbola, 2 2 2 sin 6 is o, that is according as according as cos 6 tan v
>
<
tan
#<=:>
tan
v.
2. For the cylinder
of revolution about the axis of
z,
we make
x=x
cos 0,y
is a2
y
cos
2
in
its
2
equation x*+y*
= r*;
/,
the curve of in
tersection
04-/
= r* an
ellipse.
EXAMPLES (Coordinates Rectangular).

1.
Find the right
line generators of the hyperboloid
*+>_:=, i
9
4
for the point (2, 3 ?)
2.
on the on the
surface.
2
2
Find the right
line generators of the paraboloid
4jy
2$z
= icxxr
:
for the point (? 2.

3.
i)
surface.
Find the planes of circular sections of the following surfaces

36
(0
(2)
(3) (4)
=i44
i2
4.
In the hyperboloid of revolution of one sheet

the
^ --- = g f r
2
2
-f. j;
find
equations of the generating line whose projection on the

is
plane xz
5.
tangent to hyperbolic section in that plane at

2
its
vertex.
Find the sections of the cone x*+y*=(z 2) by planes con taining the axis of j/, at angles to the plane xy of 30, 45, and 60
respectively.
6.
Find the curve of intersection of the surface
by a plane inclined at an angle of 30 to the plane xy, and whose trace on that plane makes an angle of 45 with the axis O.v.
CHAPTER
X.
TANGENT PLANES, DIAMETRAL PLANES, AND

CONJUGATE DIAMETERS.
123.
line
Straight
order. meeting surfaces of second
We
can transform the general equation
to polar coordinates
by writing
/r,
mr
nr,
(when
/,
m, n
are in rectangular coordinates, direction cosines,

ordinates, direction ratios).
r*( A/
2 2
2
and The equation becomes
in oblique co
+ Bfl* + Cri + 2K ?nn + 2 Win + 2C7m) + F= +C"
o.
Hence
two points be coincident the line

124.
a straight line meets the surface in two points, and is tangent to the surface.
to
if
these
Tangent Plane
surfaces
of second
order.
Let the origin be on the surface (and therefore o) then one o. of the values of r in (2) is r Now, in order that the radius vector shall touch the surface at the origin, the second root must be o. o, and the condition for this is A"/+B"w + C"n Multiplying
F=
this
by r and replacing
Ir,
mr, nr by x,
j>,
z,
this
becomes
= o.
Hence
fixed plane (3);
(3)
the radius vector touching the surface at the origin lies in the
and
the locus of
origin,
all
c is n are arbitrary, A",r + B j + C"2 touch the surface at the the radii vectores which
as
/,
;;/,
and
is
Hence,
if
the
therefore the tangent plane at the origin. equation of the surface can be written in the form
u.2
+ !=
o (where
represents terms of second degree
and u terms
{
76

of first degree in #, y, and gent plane at the origin.
z),
then u
is
the equation of the tan
Therefore, to find the equation point x y z transfer the origin

,
to
the tangent plane to the surface at the

this point.
is
to
writ/en
u2
-f Uj
= o,
and
Uj
;
=o
l he equation may then

to
be
the
tangent plane referred

retransfer the origin
the
point of contact as origin

primitive one.
then in
Uj= o
to the
125.
For the
central
surfaces (origin at centre) take the equation
and
let
(x ,y
,
z
r
,
be the point of contact.

z
)
Transferring the origin to
the point (x
by the formulae
= X+X
1
Ax* + B/ + Os 8
y y +y V 2= Z + Z + 2 Axx + 2 fyy + 2Czz =o.

}
we have
Hence
the tangent plane at the
new
origin
(
is
Axx
-f
By/ + Czz
= o.
Now
formulae
retransfer the
origin for equation (i)
to
the centre by the
*=*-*
+ Czz
or AAVT
plane, at the point
=x x =y y
Ax
*
\
)
2
and we obtain
Bi/
CV
o,
+ BJ/ + Czz =i xy z
(2)
the required equation
of the tangent
referred to centre.
i. For the sphere A
=B = C =
-^ a
Hence
(2) gives
xx +yy + zz
= a\
B
(3)
2. For
the ellipsoid
A=
xx
-i
-,
=
zz
-1
C=
yy
3. For
the hyperboloid of two sheets
A=
~ E~
^ C=
xx
7*
yy
zz
78
.VOTES
4. For
ON SOLID GEOMETRY.
one
sheet,
the hyperboloid of
A=
B=:
C=
^.
xx
yv
126.
treating the
For the surfaces which have no centre (origin at vertex) by 2 equation By + Cz* = 2A. x in a similar manner we
(x + x (x + x
(7) for the equation to the tangent plane
obtain
Bj/y
+ Oss = K
Czz
to the elliptical paraboloid
and
(8) for the tangent plane to the
Bij/
= A.
hyper
bolic paraboloid.
Remark.
linders.
The same method may be
applied to cones and cy
127. Polar planes
to
surfaces
(4) (5) (6) (7) (8) are the
of second order. The equations (3) equations to the polar planes to the sur
f
faces respectively with respect to the point
y z ) and these polar planes possess properties analogous to the polar lines to the conic
(.v
, ,
sections.
128.
The length of
the perpendicular
from
the centre on the tangent

2
plane to the ellipsoid is cos a, cos (3, cos y are
p
its
\Az cos
a + &* cos
ft
+ c* cos y
2
when
direction cosines.
The equation
also be written
to the tangent plane
is
"-^
"-jrjH
g-
i.
It
may
cos
+y cos
cos
z
ft
z cos
p.
Hence we must
have
/~
i
cos
cos
/?
a cos
_ b cos
ft
_ c cos y _
_s/
y
^~
x
~a~
y
~T
~J
~^~
~7
Hence
calling the direction cosines
/,
z,
w,
the equation of the
tangent plane
may
Ix
4-
be written
my -f
.s
= V^ + ^ ^
2
+^
(9 )

129.
79
To find
the condition thai the plane

3C
a
x*
the ellipsoid
f-
-~
ft
\f
2
H
=i
r
(i) shall be tangent
to
y
z*
-\
a
(i) with
5-
~-r
u
xx
Comparing
^
+ -^
yy
22
-\
we must have
I
"
X
o~ 2
j
y
/o
i
a
a
or
7T~
ft
6*
y
c
x
a
,
^f
r
ft
= -^-,
y
by
r
=
is
z
;
/.
squaring and
adding
<^
+ ^
p
the
required condi-
tion.
130.
centre
The sum of the squares of the perpendiculars

the ellipsoid
\>,
\>"
from
the
is
of
on three tangent planes mutually at right angles
constant.
Let cos
cosines.
cos
ft
cos y\ cos
<*
cos
ft
cos j/,
etc.,
be the direction
Then
p p
= = =
cos cos
-f
2
<$
cos
/5 /?
ft"
+<;
cos cos
a"
a +b~
a"
cos
cos
+t
}-
y y
,
2
<z
cos
+^
cos T/
and adding we have
131.
planes
to the ellipsoid
Cor. Hence the locus of the point of intersection of three tangent which intersect at right angles is a concentric sphere
of the radius V + b* + c For 2 the square of its distance from the centre p*+p *+p and therefore to a* + tf + c\
1
.
"\
is
equal to
Remark.
a2
2
.
b~,
is
negative,
In the case of hyperboloids one at least of the quantities and hence their sum may be negative or nothing
;
through which three the hyperboloid can be drawn, and in rectangular planes touching the latter case the centre is the only point which has that property.
in the
former case there
is
no point
in space
8o
NOTES OA SOLID GEOMETRY.
A diametral surface is the Definition. 132. Diametral Planes. locus of the middle points of a series of parallel chords of a given surface. Diametral lines or diameters are the intersections of the
diametral surfaces.
133.
To find
the
diametral surface corresponding
to
a given series of
centre.
parallel chords in
a surface of the second order which has a
Let the equation of the surface be
/,
m, n the direction cosines of each the parallel chords, and
/, z
the coordinates of
middle point. The equation of the chord will be

its
x
I
oc
z
=:
zn
i*.
n
it
Then
have
for the
points in which
meets the surface (i) we
shall
or
Imposing on this the condition of equal roots for r, we have Atx + Bniy + Cnz = o (2) the equation of the diametral surface, a
plane passing through the centre.
134.
The diameter
/
=
n
is
one of the
series
of parallel
chords bisected by the plane

gate to the plane,
luxate to the diameter

If a diametral plane
and is called the diameter conju (2), o is conand conversely the plane lx + my + nz A.v Cz Bv -
-.
be chosen as a new plane of xy and its con jugate diameter be taken as the new axis of z, the centre O being still the origin; then, since every chord parallel to Oz is bisected by
the plane xy, the equation of surface will contain only the second power of z. Hence, if there be three planes through the centre the
intersection of
any two of which
is
conjugate to the third, the equa
tion of the surface referred to these planes will be of the form
AV + By + CV=i,
that
is
(3)
of the same form as the equation referred to rectangular axes.

!35-
81
the centre
To find
the conditions that
of three planes through

be
of a
surface
of the second order each
may
diametral
to the intersection
of the
other two.
Let the planes be

/v
_|_
mv + nz
= 0,
x + my + n z
= o,
l"x
4-
m y+
!
n"z
o.
The equations
of the diameters conjugate to the
first
plane are
and
if this
be parallel to the other two planes, we shall have

/
+M
I n m oandr + + A B "c=
,
.
...
,,
"
m
B+"
,,
c= 0;
o,
these with the third equation

like
+m
found
in
manner, are the required conditions. These three planes are called conjugate planes, and
their intersec
tions conjugate diameters. Since we have only three relations

will
be an infinite
number
between the six quantities there of systems of conjugate planes in each

If in (3) Art. 134
surface.
136. Equations referred to conjugate diameters.
we
make
A _
>_
i B a v
,
_JLv ,
- -& c-^
Then
X~
for the ellipsoid

2
I
2
-f
will
c
be the equation referred
to
conjugate di
ameters, and a, b
will
For the hyperboloids we

a2
be the semi-conjugate diameters. shall have
^-^- 7T=
z*
and
jr^-^bc
r.
Remark. The tangent planes at the extremities (.* j/, z) of any diameter to a central surface are parallel to the diametral plane conjugate to the diameter so that the conjugate plane of the diameter
, 1
through the point (x
on the
vy
ellipsoid
is
xx
zz
82
137.

The sum of the squares of
three conjugate semi-diameters
of the
ellipsoid is constant.
In the first place, any point on the ellipsoid may be represented by the equations x =. a cos A, y b cos //, z c cos Y, when cos A, cos ju, cos v are the direction cosines of some line, for the condition 2 cos A + cos 2 //-fcos 2 v i cause these three equations to satisfy the
equation of the ellipsoid. Therefore if cos A, cos fa cos r, cos A cos //, cos v are the direc tion cosines of two lines answering to the extremities of two conju
,
gate diameters, these will be at right angles to each other.

T,
ror the equation
xx
yy +~j^yu
zz
g-= o
will give
cos A cos A
+ cos
cos //4-cos
v cos v
o.
2
the square of the length of any semi-diameter expressed in terms of A, /i, v, is

2 rt
Now
~v
+y +V
2
= a* cos
in
2
A+
2
<
cos
,
2
JJL
+ C*
v\
cos
r,
and of the conjugates

2
terms of A
2
//,
2
A",
//
2
y"
=rt cos A
a* cos
2
cos
2
yu
2
<:"*=
A"+3
cos
+r // +
cos
r
2
v".
r>
cos
Adding we have
2
a"
+ 3 /2 + c
y"
= a + tf +
2
2
<r
since the
lines
A,
ju,
r,
//
r x and
,
A",
yu",
are mutually at right angles.

the locus
138.
To find
of the
intersection
of
three tangent planes at the
extremities
of three conjugate diameters.

of the three tangent planes are
The equations
a
OC
cos A
+v o +
-f
cos
V
u, -f
- cos v
c
cos A
(2
cos
O
v
V
,,
-\-
- cos v
C
=
V
x a
COS A
y
b
z
-\
COS
COS
c
Squaring and adding, we get

H
>
for the
equation of the locus

,
~-f--r =
an ellipsoid with the semi-axes a\/~$~, b^/ 3
tV~T-

139.
83
The parallelepiped whose edges are three conjugate semi-diameters

a constant volume.
a, b, c\
,
of an
r
ellipsoid lias
Let Qx, Oy, O.2 be the semi-axes of the surface Oz any system of semi-conjugate diameters a b
,
Oo;
Oy,
c\ let the plane
of
Oy
x y intersect that of xy in the = B be the semi-diameter of the

Hence parallelogram a
.-.
semi-diameter OjCi
curve
CXrj.
=
f
)
^#
which
is
A, and let conjugate to
parallelogram AB.
Vol (a, b\
= Vol
(A,
B,-O
have the same altitudes and equal bases. Let the plane z Oj 2 intersect xy in the semi-diameter Oi^ C, then this plane must contain Oz for, being conjugate to OA*, in a principal plane it must be perpendicular to that plane hence CXv,,
for these figures
;
;
Or,.
Oz form
them
a system of semi-conjugate diameters, and any two of are semi-conjugate diameters of the plane section in which they
.-.
are situated.
Vol (A, B, Vol (A, C,
=Vol
=Vol =Vol
(A, C, c)
(a, b,-c) (a,
6,
c)
c
)
Vol
(a
c).
140. To find
the diametral plane bisecting a given system
ofparallel
chords in the case of the surfaces which have not a centre.
Taking the equation of the surface
and one
or the
chords , ,
xx
I
;;/
= vym
z =. ~
z -
=r
the equation of the diametral plane will be

BJ/
+ W Cz= A
7.
Hence
the diametral planes are parallel to the
common
axis of the
principal parabolic sections.
cannot, therefore, in these surfaces have a system of three con jugate planes at a finite distance, but we can find an infinite number such that for two of them each bisects the chords parallel to the other
We
and
by proceeding as in Art. (135). the origin where the intersection of these two meets the paraboloid, and referring to these three planes, the equation of the surface will be of the form
to a third plane,
By taking
84
NOTES
the third plane
origin.
is
OiV
SOLID GEOMETRY.
And
the
evidently the tangent plane to the surface at
new
141.
The tangent planes
to the
hyperboloid of one sheet
and
/he hyperbolic
paraboloid at a point x y z
intersect the surfaces each in
two right
line
generators through the point of contact.
The equation
jugate diameters
of the hyperboloid of one sheet referred to any con

is
and the equation of the section made by any plane

to the conjugate plane of xz,
is
ft parallel
and
it
is
evident that the value
/3b
(.v
,
gives us the section of the

,
tangent plane at the extremity
of the diameter b
or
x
Ct
3
z"*
~~^
c
7?
>
two right line generators.
For the hyperbolic paraboloid

8
B>
-CV=2E".,v
is
(i)
the tangent plane through the origin with (i) is
o,
and
its
intersection
B^
CV=
o,
two right
line generators.
CHAPTER XL
GENERAL EQUATION OF THE SECOND DEGREE x,y, AND
z.
IN
142.
eral
In order to discover
all
the surfaces represented
by the gen
numerical equation
Bj/
Gz
+2C ^ +
2C"a
=D
(E).
origin by
we
will first
transform the coordinates to a
new
means of
the formulae
x
z
=.
a + x*
and endeavor
origin in
to
such manner as
If this
determine the coordinates (a, to cause the terms of the

will
/?,
y) of the new
first
appear.
can be effected the equation

2
degree to dis be reduced to the
form
Ax
in
4-
Br + Cs + 2A. zy + 2%
zx
+ 2C xy = F
(F)
is no change when z are substituted for x, y, +z, and which therefore represents a surface having a +y, centre, and the new origin of coordinates is at this centre.
which there
+ x,
Now,
relations
several different cases
may
arise
according to the numerical
among
ft,
the coefficients A, B, C,
x
,
A",
B",
C".
i.
a,
the coordinates of the centre
may
each have a
finite
value found from the three equations determining the conditions of the transformation.
2.
a, a,
/3,
y may
y may
8
have
infinite values.
3.
/?,
be indeterminate.
85
86
A 07ÊS
T
ON SOLID GEOMETRY.
The
surfaces corresponding to these three cases will be
(A) Surfaces having a centre.

(B) Surfaces having no centre (centre at an infinite distance). (C) Surfaces having an indefinite number of centres.
143.
Making
the actual transformation of (E) by the formulae (i
we have
And
appear,
in order that the terms of the
first
degree in x, y, z shall dis
we must have
(C)
which are called the equations of the centre.
1. If these three equations give finite values for a. /?, y, then the surface represented by the given equation has a centre.
2.
If
two of ihese equations are incompatible

/3,
this
shows
infinite
values for a,
y, and the surface has no
centre.
3.
plane,
is
If the
line of centres.
and
three equations reduce to two, then the surface has a For each one of the equations is the equation of a two taken simultaneously represent a line, and the surface or hyperbolic cylinder.
For, cut the surface by the
an
elliptical
and Q containing planes P and Q, P cutting the line of centres (D) The section by P is a curve of the second degree having that line. The its centre on the line D, and hence an ellipse or hyperbola.
section
will
be two straight lines
parallel to the line
D, and as
sec positions giving two straight line may tions parallel to D, the surface is a cylinder.
revolve about
in all
its
surface equations reduce to a single one, then the coincident has a plane of centres (i. e., the given equation represents
4.
If the three
or parallel planes).
Note.
The equations
of the centre can be found
in
any given

and
87
equation most readily by finding the derived equations with regard

to x,
y,
z respectively
,
(i.e.,
by differentiating with regard to x. y, z
respectively), the x, j
and
2 in the resulting equations standing for
a,
ft,
y.
144.
Example
,v
2
i.
Determine the
1
class
of the surface represented

2
by the equation
+ ^y + 42 + 2yz + \zx + 6.17

2
6x
2$y
322= 26.
The equations
of the centre are
6>
+ 22 + 6x
24
82 + 27 + 4^32
and the surface has a
centre.
=O = 0)
>
These give
y =2
*=3J
Example
2.
Determine the
class of the surface
The equations
of the centre are
+ 22+2 V 4 = y + 22 + 2X 2 = 4 Z + 2y + 2X + 2 =
2JI-
O
O
O
)
f x
,
the
first
two of which
x+y+z =
2,
x+y + z =
infinite,
are incompatible,
hence the coordinates of the centre are no centre.
and the surface has
Example
3.
Determine the
2
2
2yz
zx-\-^xy+2z
o.
The equations
of the centre are
2x
8j/
+4y ~ 22 + 4^v =
2V
o
o
2Z
X+ 2=OJ
}
The
duce
to
first
two of these are identical, hence the three equations re two and the surface has a line of centres (/. e. is a cylinder).
4.
2
Example
8-v
2
Determine the
2 1
8v
+ 22 +
272
+ 8zx + 2^xy +
82
5o.v
757
252
+ 75 =
o.
The equations
of the centre are

1
6x
+ 247
8.v
50
i2r+
25
= = =
o o
88
which are
all
NOTES OX SOLID GEOMETRY.

three the same, each being 8.r+ 4-4.3 =25. Hence and consists of a pair of parallel
i2>
the surface has a plane of centres,

planes.
145. Recurring to the general equations of the centre
<*4-C7?
+ By + A"= o + ir=o(
]
(c)
an easy rule for a relation among the coefficients in any given equation by which we can distinguish the central surfaces from those having no centre and those having an infinity of centres.
find
we may
The common denominator

equations
is
of the values of a,
/3,
and y
in these
the determinant
B A, C C B, A B A C
,
+ CC -ABC-2A B C
2
Now,
if
R=
it
may
be different from zero, the surface has a centre either have no centre or an infinity of centres.
but
if
The
form
value of
A, B,
R may
be written out by the following
mnemonic
for the first
A B C
A B C
the letters to be multiplied by three terms,
columns
two
and by rows
the
for the
last.
146.
To find an easy
equation
rule
for F,
formed
centre.
of
the central surfaces
new absolute term in the trans when the origin is moved to the
is
This complete transformed equation

A_v
s 2
+ By + Or + 2 A zy + zR zx + 2C xy = F when
F=DNow, multiplying the first of the equations (C) of the centre by and the third by y, and adding them ft,
a, the second by
we have

Hence F = D
substitute
(A"
89
+
(i.
B"/?
+C
|-/?,
Therefore the rule for
is
for
x,
y,
in the
e.,
terms of the first degree one-half the
co
ordinates
of
the
centre
â,
^y
respectively
and
take
result
from D. Example we have
i.
Taking the Example
i,
Art. (144), in
2,
which the coordi
nates of ihe centre are found to be
x =-i,y
i
= 3,
1;
F=26 + 26
X* +
2
J-
+ 24 x +32
is
xf
1 1
and the transformed equation

2
3>
= + 43 + 2yz + 4zx -f Ex. 2. 2^ + 3y* + 4Z + Syz + 6xz + 4xv 6x SyI4z = = 2, s= Here the coordinates of the centre are x =
6,r>
1 1 1
20.
i.
\>y
F==2o +
6 x-J
14 x
= 17;
1
and the transformed equation

2 A* 42
is
3/ + 4-s + 870 + 6^2- + ^xy
7.
147.
Removal of
the terms in xy, xz, yz.
Reduction of the equation of
the second degree to
two forms.
Fora more complete discrimination of the surfaces represented by the general equation, we will now remove the terms in^ry, xz, yz by So far we have made no supposition a transformation of coordinates.
as to the direction of the axes.
Henceforth, for convenience, we will
consider the axes rectangular.
Taking
transform
the equation (E) in rectangular axes we propose now to also rectangular in such manner that the it to a system
terms in xy, xz, yz shall disappear. The disappearance of these terms can only be effected by taking for coordinate planes either dia metral planes or planes parallel to them.
We
148.
will
therefore begin
by finding a diametral plane conjugate
to
a given diameter.
To find a diametral plane conjugate
to
a given diameter.
x
I
yy
m
y
to
r,
__
zz
n
-\-nr in the general
Putting
first
x-= x + lr, y
in r
+ mr, z=z
equation,
and arranging with reference
we have
for the coefficient of the
deree
QO
and
this

placed equal to zero
is
the equation of the diametral plane,
namely
(
A/+ B
149.
+ C m)x + (C7+ Bm + A n}v + (B7+ A m + Cn)z + A 7

70 determine a diametral plane perpendicular
that
is,
to the
chords which
it bisects,
to
find a principal plane.
In order that the diametral plane shall be perpendicular to the line
m
I
we must have
the conditions fulfilled
m
s.
or putting each of these equal to
At
+R n+C m =
(A)
and
also the condition /
+w
-f
#2
To
determine
/,
w, and n
in
equations (A) we
first
find
s.
Writ
ing these equations
they give the result
A-J, C B C B j, A B A C-.r
,
,
=o
or
or
C *-ABC-2A B C
This cubic has necessarily one real value for (A) gives one set of real values for /, m, n.
s,
=o
is
(D).
which substituted
in
Hence
there
one
principal plane.
For convenience of discussion
let
us take this plane perpendicular

to the axis of
z,
gi
then
= o, w = o,
o,
tions (A) give
o,
A=
to this principal
plane as
and n =i. And hence equa and the general equation transformed plane of xy is of the form
z
D.
Now we know
transformation
is
from the like discussions in conic sections that one
always possible, and but one to a system of rectan gular axes in the plane xy which shall cause the term in xy to dis Hence there are three principal planes, and three sets of appear.
values for
/,
m,
n,
and the cubic (D) has
three real roots.
general equation coordinates to the form
The
may
then be always reduced in rectangular
D.
(E
which represents then

1
all
the surfaces of the second order.

this equation
50.
+ 2N z
=D
The reduction of
to
Lx 2 + My 2 -f Nz + 2L
2
+ 2M y
two forms.
i. If L, M, and N are different from o. Then we may cause the terms of the first degree
transferring the origin to the point
to disappear
by
x
z
)
L
~r~>-J
1 4
M
ivf
i* L
N
*
IN
\f
The
will
surface will then have (#
,.>
for its centre,
and the equation
be of the form
Lr + My + Ns ^
2
F.
(I.)
2.
and
If
one of the three

be different from
coefficients, L,
o.
M, N,
for
example
We
cannot then cause the term 2~L!x to disappear, but by trans
ferring the origin to the point
D = --., y = 2L
M M
= --N
f
^pN
the equation will take the form
M/ + Ns =
2V x.
(II.)*
The forms I. and II., we have seen, belong to the surfaces of the second order, which we have already discussed. Hence the general equation of the second degree (E) represents these surfaces and no
others.
QO
151.

The form
I. we have seen represents the ellipsoid, the two and cones of second degree, and includes the elliptic hyperboloids F and parallel planes N2 2 =F. and hyperbolic cylinder, Mv 2 + Ns 2
-
The form
and
II.
represents the elliptic
and hyperbolic paraboloids,
the parabolic cylinder.
152.
The complete reduction of

I.
the equation
of the second
degree to the
Use of the discriminating cubic (D). simple forms The resolution of the equations (A) furnishes for each value of s
II.
and
in the cubic (D),
one system of values

/
,
of/, m, n.
We have
then three
which are the direction cosines systems, /, , of the three rectangular axes {principal axes ] to which the surface
m, n]
??i
,
/",
m",
must be
appear
;
referred
in
order to cause the products
.%T,
xz,
yz
to dis
the formulae of transformation are then
y
z
If
=
2
mx + my + nx + n y +
x*
2
m"z
n"z
we take only
the terms in
2
in this substitution
we
find
L
But
if
= A/ + Bw + C;* + 2h. mn + 2BW+ zC lm.

the equations (A) respectively by
/
2
/,
we multiply
that
m, n and add,
remembering
+
2
2
7
7z
we have
s
;
A/ 2 +
B;;;
+ CV + 2h. mn + 2 BW+ 2C lm
Hence L
roots.
is
a root of the cubic (D) and
M
have
and
are the other
two
For the values of
M N
,
we
r
will
=A7 M=A7
D
+B"w +C"n
+B";;/
+C
(M).
The
absolute term
is
does not change in this transformation since
the origin
not changed thereby.
For the surfaces having a single centre after solving the cubic, we have only to calculate F, for which we have given a rule.
For the surfaces having no centre the coefficient designated by V L and is computed by first finding in equations (A) is equal to Both in the cases o. the values of/, m, n, which correspond to s
,
one root of cubic of surfaces having no centre and a quadratic to solve to determine L and M. o and we have only
a line of centres,

153.
93
For surfaces having a
centre,
particular
class of the surface,
tell us whether the surface is an ellip one sheet, or hyperboloid of two sheets. These * without signs we can ascertain from inspection by Descartes s rule
formation of the equation to its of the discriminating cubic will
if \ve wish only to discover the without making the complete trans centre and axis, the sign of the roots
soid, hyperboloid of
solving the equation.

2 2 2 Example. Find the nature of the surface 7# + 6y +5.s
4_>
$xy
or
6.
The
s
3
cubic (D) gives
(7
+ 6 + 5)s + (42 + 35 + 30 -4 4)5+28 + 20 i8s* + 99^162=0. s

a 3
210=0;
.-.
The row
of signs
is
three changes of sign.

is
Hence
all
the roots are
and the surface
an
ellipsoid.
So also for surfaces having a line of centres, the signs of the roots of the quadratic into which the discriminating cubic degenerates, serve to distinguish the elliptic from the hyperbolic cylinder.
And for surfaces having no centre, the signs of the roots distinguish the elliptic paraboloid from the hyperbolic paraboloid.
154. Recapitulation of the method of reduction of numerical equations of the second degree and of distinguishing the surfaces represented by them.
now propose to give the mode of distinguishing the nature of the surface represented by any given numerical equation of the second degree in x, y, and z, and of finding its principal elements.
I. Form the equations of the centre, and also the discriminating cubic from the remembered form
We
s *-
z (A + B + C)s +
AB + AC + BC - A - B -C + AA + BB + CC -ABC-2A B C =
2
2
).r
o,
observing that the absolute term is equal to R, the denominator of the values of the coordinates of the centre in the general equation,
ABC
and therefore can be formed by the mnemonic
ABC ABC
(Art.
1
145).
Then
155.
i.
If
be different from
o,
the surface has a centre.
Find
Note.
changes of sign
number
the roots being real the number of positive roots is equal to the number of row of signs of the terms, and the number of negative roots is equal to the of continuations of sign."
"All
in the
94

Determine the signs of the roots of the cubic by Descalling these roots L, M, and N, and calling F
the coordinates of the centre and transform to the centre by the rule
in Art. (146). cartes s rule.
Then
the
new
absolute term on the second side of the equation.
Then
a.
If L,
M,
N N
all
have the same sign as F, the surface
is
an
ellip
soid.
b.
If L,
M,
all
have a different sign from F, the surface
is
im
aginary.
c.
If
surface
d.
two only of the roots L, M, N have the same sign as F, the is the hyperboloid of one sheet.
If only
is
one of the roots L, M,

o and L, M,
has the same sign as F, the
surface
e.
the hyperboloid of two sheets.
If
F
F
all
have the same sign, the locus
is
point.
f.
If
=o
If
and one of the roots L, M,

is
N
its
has a different sign

(Art. 85).
from the other two, the surface

156.
an
elliptic
cone
2.
R=
2
,
o the cubic has one of

s,
roots s
o and
is
degraded to a quadratic, the coefficient of
namely
AB + AC + BC
B
if
becomes the absolute term.
And
the equations of the centre are incompatible the surface has
no centre.
Then
a.
If the roots
(i.
M
e.)
and
if
the
is
same sign
If
2
AB + AC + BC
of the quadratic (degenerate cubic) have 2 the surface A2 B
*C
>o
the elliptical paraboloid.

b.
M
C C
2
and
have different signs

is
(/.
e.)
if
AB + AC + BC-A"
B
c.
*<o
the surface
the hyperbolic paraboloid.
If
2
one of the roots
or
is
be zero
(*
e.)
if
AB + AC + BC-A
=o
the surface
the parabolic cylinder.
to
3. If R = o and the equations of the centre can be reduced two equations, the surface has a line of centres. The cubic as in (2) has one of its roots S = o and degenerates into the quadratic
157.
j_(A+B + C)s + AB + AC + BC --A"-B

Then
a.
-C"
= o.
(i.
If the roots
and
of this quadratic have the same sign
e.)

if
95
an
elliptic
AB + AC + BC
b.
A"
~>
o the
surface
is
cy
linder.
If the roots
M
C
2
+ BC
c.
"
B
and
<
and N have different signs (i. e.) if AB + AC o the surface is the hyperbolic cylinder.
If in the reduced equation of the cylinder

o,
}
Mz
-f
Ny 9
=
is
H,
be
equal to
z
line
and
both of same sign, the locus
a straight
=o J=o d. If H =
>
o and
and
be of different signs the surface con
sists
of intersecting planes.
o and the equations of the centre become a and consists of two parallel or coincident planes, which are readily found by solving the equation with reference to any one of the variables.
158.
4.
If
R =
single equation, the surface has a plane of centres,
159.
roots.
5. In the case of surfaces of revolution the cubic has equal
To examine
surfaces of revolution,
the cubic for equal roots in the case of central we simply look for a commoe root between it
and
1
its first
derived equation (differential).
60.
GENERAL REMARK.
In any of the above cases we
may com
plete the reduction by solving the cubic to get the new axes and thus obtain their direction by finding /, m, n from equations (A). And in the case of the surfaces without a centre we may find V,
from equations (M).

161.
REMARK
I.
and of those not having a

the coordinate planes.
In the cases of surfaces having a line of centres centre, we can distinguish readily the sur
face represented by a given
numerical equation through sections by
i.
If the
by the coordinate planes

a hyperbolic cylinder.
equations of the centre show a line of centres, sections will tell whether the surface is an elliptic or
2. When
a.
the equations of the centre are ellipses

is
show no
centre, then
If there
among
these sections by the coordinate
planes, the surface

b.
an
elliptical
paraboloid.
these sections, the surface
is
If there are hyperbolas
among
hyperbolic paraboloid.
c.
If all
these sections are

is
straight lines, the surface
parabolas, or one of a paraboMc cylinder.
them
parallel
96
162.

REMARK
II.
Again,
if
the terms of the second degree in the
given equation break up into unequal real factors, the surface must be either the hyperbolic paraboloid or hyperbolic cylinder, and these two surfaces are otherwise readily distinguished. We may note also
that if the terms
perfect
square, the surface
of the second degree in the given equation form a either a parabolic cylinder or two is
parallel planes.
163.
We
i.
will
2
now
illustrate
i
by a few examples
Ex.
yjt-
iy~ + 6z + 2^xy+ \2yz- 1202:

is
84.
As
this
is
a central surface with the origin at the centre,
we only
need the discriminating cubic, which

j
8
343^4-2058
or
s*os*
343^4-2058
= o.
signs 4hence the surface
The
-is
show one continuation and two changes, and 1a hyperboloid of one sheet, or two sheets, accord
is
ing to the sign of 84.
By
fore
ing the equation
and then by depress 21. There two roots are 14 and the equation of the surface referred to its centre and axes is
trial
we
find that 7
a root of the cubic,
we
find the other
7*
Ex.
2.
-f
14^-212*=
2
84; or.r
2
+ 2_y -3S =:i2.

2 2
2 5.r
+ 2 2_/ + i6s
4-
i6yz4zx
2oxy
z6x
40^442
= -46.
The equations
of the centre are

250,
I
ioy
22
= =
13
o:r 4- 2 2j 4-
82=20
162
22
;
2X+ 8y+
whence we
find the coordinates of the centre
\,y =i, * =i.
Moreover
F
The
=
/
46
4-
26
is
\ + 40
J + 44
9-
discriminating cubic
63^+ 134^5832
1
o.
Its
The
signs give three changes. surface then is an ellipsoid.
Hence
all
the roots are positive.

By
trial
97
we
find that 9
is
one of the roots of the cubic.
Hence
the
other two are 18 and 36.

J
The reduced equation

2
is
then
2
-fi8>
9jr
-f
36s
or
A-
And
the principal semi-axes are
__
V2
1
Ex.
3.
5 x*
+ of +
i
ys
+ 2 6yz +
8zx + i $xy + 6x +
?>
02
= 64.
The equations
of the centre are
5.r+
7^+
92
7A:+iqy+
132
=4
i
=-3
Multiplying the
first
of these equations by
the third.
and the second
by
2,
and adding, we obtain
Hence
only two independent ones. The surface is tersecting it by the coordinate plane xy, i. e.,
2
the equations are In therefore a cylinder.
making
64,
o,
we obtain
5^
which
+ 1 4xy +
oy*
+ 6x + Sy
To
The surface is therefore an elliptic cylinder. is an ellipse. complete the reduction we transfer the origin to the point
line of centres
>
where the
x, y, that
pierces the plane
is,
to the point z
o,
y
is
i,
2,
and
find
F=64 +6
4=:66.
Also the discriminating cubic

3
.r
32,r-f6 9
>
o,
which gives
5\/
is
32^ + 6
o,
the roots of which are 16
+ 5^
10 and 16
10.
And
the reduced equation of the cylinder
io)jr
+ (i6
5V
7
io)>
=66.
30
Ex. 4
2
.
5 jp
5/
+ 8s + 4sv + \zx
2
Sxy + 6x + 6y
o.
The equations
of the centre are
^4-4*=
98
Adding
first.

the two last of these equations
2j.
we have
is
$x4y+2z
Hence
An
2
iS>r
equation which
incompatible with the
the surface has
no
centre.
The cubic
+ 8ij=
to 9.
or
s*i 8^ + 81 = o,
surface
is
which gives two roots equal

boloid of revolution
The
2
therefore a para
To
find
V, we
first
determine
4/
/,
and
272
n.
For these we have the
equations
=
o
U- U 7 which Rive /
2
,
n
3
-!
Therefore
(Eq.
M)
L
2V
= 3 .-+ 3 .-+-.- = 2L
and
=9.
is
The reduced equation
of the surface
therefore
Ex.
5.
2A70+2B ^v+2C .r^

is
4-2A
v
.v
+ 2B
jv4-2C"2;
D.
The cubic
s-(A + B
2
f-C
if
)s-2A B C
,
=o
The
o.
surface
is
is
a hyperboloid
A B
and C
are all different from
of the same sign as F in the reduced equation the cubic will have two roots of the same sign as F and the surface will be a hyperboloid of one sheet. In the opposite case it would be a hyper
If
boloid
If
2
j.
A BC
A =
_(B
of two
2
sheets.
o the cubic becomes
4-C
*)
=o, whose
roots are of different signs.
Hence
the surface
2H zx + 2C xy-\2 2
2A".r+
2B V + 2C
is
a hyperbolic
paraboloid.
Ex.
6.
jc
+_y 4-9s
+ 6r0
6xz
2~\y+
2.v
42
o.
The equations
of the centre are incompatible and the terms of the
O.Y
SOLID GEOMETRY.
is
99
a parabolic
second degree form a perfect square, hence the surface

cylinder.
EXAMPLES.
164.
i.
Find the nature of the surfaces represented by the follow

2
ing equations.
= o. 2o\ z + 4zx 4- 6xv 4- 2 2.v + 6 v 4- 42 4+ 2X2 4- 2AJ O,V OV IOZ-\-2$ = O. +/ ( 6x + 3^ = 0. 3 jc 3 37 2J2T +i2zx + Sxv 60jt- + 54^ =36. 4.i- + QV + 97s (4). 8 = O. 80 4VZ 4.V (5). 3X*+2f2XZ nzx jxy = ^ represents 2. The equation 7^- 4-8/ + 4s 7^0
(
i
).
i~v -f 5_y 42
2z
).
.X
4- S* 42
2F
).
6>>
\-
a hyperboloid of one sheet.

1 d 12 +32: + yz+zx+xy 252 jx represents an ellipsoid, a point, or an imaginary surface according
3.
The equation
<
.r
;2
+>
4>
as
d
4.
is
>
67.
_v
2
The equation
4-.v
4-2 -i-yz+zx
+ xy =
z)*+
\
a* represents
an oblate
2
spheroid.
5.
Find the nature of the surface (y
(z
xY + (x
#
2
.
v)
a~.
6.
Find the nature of the surface yz

â2
zx + xy
o re 7. 4-9s presents an elliptic, a parabolic or a hyperbolic cylinder according as

4-
+ 4v
i.
I2yz
+ 6zx + 4xy+i4x+ 16^+2404-47 =
>
<
CHAPTER
XII.
PROBLEMS OF
165. PROB.
LOCI.
line
I. To find the surface of revolution generated by a right turning around a fixed axis which it does not intersect. Let the fixed line be the axis of z and let the shortest distance a
from the revolving line to the axis of z

original position of this line so that
its is
lie
along the axis of

is
equation
a,
x y=
in the
mz.
Then
the equation of the surface
or
The hyperboloid
of revolution of one sheet.
Prob. 2. To find the locus of a point whose shortest distances from two given non-intersecting, non-parallel straight lines are equal. Take the axis of z along the shortest distance between the two
lines, the
distance
2c,
plane xy perpendicular to z at the middle point of this and the axes of x and y bisecting the angles between the
projections of the line on their plane.

will
Then
c
the equation of the lines

)
be
z
=c
mx
(
y
and we have
or
cz(i
(z
cY +
y
(*
mx
mxY rL = + m~
xo
V+ ~- mxV
i
8 ///
+ ?n*)
3.
line.
-f
mxy
centre and
its
o, a hyperbolic paraboloid since it has no term of second degree breaks up into two real factors.
Prob.
straight
7 wo
planes mutually perpendicular,
contain
each a fixed
intersection.
To find the surface generated by

2.
their line
of
Take
the axes as in Prob.
Then
the equations of the planes are
100

K(z
c)
IO I
+y
mx
(i)
(z
+ t)+y + Mx
G.
(2)
The condition
of perpendicularity of these planes

o,
is
KK
+i
m*
(i)
and equations
and eliminating and (2) we have

(
i
K+K
i
between
this
equation
y _ /;/V + _ m *Y
which represents a hyperboloid of one
- m*y
sheet.
Prob. 4. To find the surface generated by a right* line which always meets three fixed right lines no two of which are in the^santy fflantj *++
For greatest simplicity take the origin at the Centre of & paraljek)-piped, and let its faces be at the distances a, &, d respectively from-
Then take three edges of this the coordinate planes yz, xz, and xy. parallelopipedon as the three fixed lines fulfilling the conditions.
Assume
for the
equations of the movable line
*^. m 2Z* cos a cos

The
(3) are respectively
-_=*.>
ft
cos
(4)
conditions that the line (4) shall meet the lines (i) (2) and
y
cos
b
ft
cos
+c y
_ ~
x +
cos
x
cos
~~~
-f
b
ft
cos
cos
Eliminate the arbitrages a, ft, y by multiplying the equations to gether, and we have for the surface
or reducing
ayz + bzx + cxy
+ abc
o,
which the discriminating cubic shows to be a hyperboloid of one The same surface will be generated by a straight line resting sheet.
on the other three edges

Prob.
x
z
=a =
}
>
y
,
=
c
}
-
= y=
x
a
.
5.
To find
9*
the
surface generated by a right line which alw ns
102

lines,
meets three fixed right
of which are parallel
to the
no two of which are same plane.
in the
same plane, but
all
Take one of
the fixed lines as the axis of x,

A:;
.
and then the other two
parallel to the plane of
Then
their equations are
Now,
-
the equations of a
.
moving
line
meeting
lines (i)
and
(2) are
this
,-
(4) (/
(.3)
and
is Ic
k arbitrary),
and the condition
that
jine shall also meet
mk (cb],
(4),
arid eliminating therV
and k by means of equation

cy
we have
mx(cb)
z b
(
or
cyz -\-m(b
c)xz
by
o,
a hyperbolic paraboloid, as its equation shows no centre, and the terms of the second degree break up into two real factors. Prob.
6.
To find
the surface generated by a right line
which meets
two fixed right
lines, always parallel Since the two fixed lines must meet the fixed plane,
"
and
is
to
a fixed plane.
we can take
and the
-y
y
(i),
[
}
(2), as in 2, as the fixed lines,
plane yz as the fixed plane.
Then
is
the equation of the

Z
moving
line parallel
toyz
arbitrary.
(3),
/,
and k
The
conditions that this line shall meet the lines (i) and (2)
mk
k+/>
-mk=-lc+t\
or or eliminating
/,
mk
k,
Ic
and/>
=
;
and
/,
mx
or
c --
y
z
mxz
= cy,
a hyperbolic paraboloid.

Prob.
7.
Ic>
Two
locus
finite
non
divided each into the same

ivhich
sion.
is
intersecting non-faralld right lines are number of equal parts ; to find the surface
the
of
the
lines
joining corresponding points of divi
lines
Let the line which joins two corresponding extremities of the given be the axis of z let the axes of x andjy be taken parallel to the
;
given lines and the plane of xy be halfway between them. lengths of the given lines be a and b.
Let the
Then
the coordinates of two corresponding points are

z =.
c,
ma,
c,
o,
mb
and the equations of the
lines joining these points are
ma
2X
+ ^-=1 mb
c
Z --- =1
?na
whence eliminating
the equation of the locus
is
a hyperbolic paraboloid.
Prob.
8.
To find
the locus
of
the middle points
of chords of a surface
of the second order that

point.
has a
centre,
which all pass through a given fixed
Take
planes which pass through

to the surface will
the given point for the origin and two conjugate diametral it for the planes of zx and xy, and a plane
of>
parallel to the third conjugate plane for that
2;
then the equation
be of the form
ax* + by 1
+ cz +
2a"x
+/
o.
Let
viz,
= nz
these with the equation of the surface,
be the equations of any chord. we have

y
c
Combining
(am*
in
+ bn^ +
)z
+ 2ct mz + d
o,
which the values of z belong to the extremities of the chord.
104
VOTES ON SOLID GEOMETRY.

its
Therefore the z of
middle point
a"m
is
and the other two coordinates of the middle point are
x = mz
Hence eliminating m and n
(2)
= nz
(3)
the required equation of the locus
a surface of the second order similar to the

its
first,
and passing through
centre and through the origin.
CHAPTER
XIII.
SOME CURVES OF DOUBLE CURVATURE.

1
66.
To find
the equations to the equable spherical spiral.
Definition.
diameter
PP
from P to
P",
meridian of a sphere revolve uniformly about its moves uniformly along the meridian while a point so as to describe an arc equal to the angle through
If a
which
the
meridian
has
revolved, the
locus
of
is
the equable
spherical spiral.
Taking PP as the axis of z, PAP the initial position of the plane of the meridian as the plane of xz, the equation of the sphere is
ac*4y +*
Let
<!*.
MON =
0,
AON
q),
then, by definition 6
=
;
cp,
and from polar coordinates
/.
Therefore
= a cos = a cos **+/ =

x
x
cos
cp,
a cos 6 sin cp
sin 0.
6, y = a cos
cos 2
(cos
+ sin*
8)
= ax.
(2)
Hence
the equations of the spiral are
.r+y +
or the spiral
If
is
(i)
#"
+ v* = ax;
circular cylinder
the curve of intersection of the sphere and a right whose diameter is the radius of the sphere.
we
subtract (2) from (i)

z~
2
<?
we obtain
(3) a parabolic cylinder.
ax
And
the equations (2) and (3) also represent the curve, which is therefore also the intersection of a right circular and right parabolic cylinder at right angles to each other.
167.
To find the equations
to
a spherical
is
ellipse.
Definition.
The
spherical ellipse
a curve traced
on the surface
105
I0 6
A OTES
ON SOLID GEOMETRY.
of a sphere such that the sum of the distances of any point on it from two fixed points on the sphere is constant. be the t\vo fixed points on the surface of the sphere Let S,
whose radius
is
;-,
which joins them.
If
the middle point of the arc of the great circle P be any point of the spherical ellipse, SP and
HP
arcs of great circles, then
SP-fHP
2a
a constant.
Through P draw PM, an arc of a SH, and let SH = 2;/, CM = y, PM

Then,
great circle perpendicular to
=
q>)
0.
in the right-angled spherical triangle
SPM
0.
we have
And
in the
= cos triangle HPM cos PH = cos

cos
SP
(y +
cos
cos
6.
J
V ?B
cos
sin Y sin A
.
a
q>
sin
/SP-HPN V
cos 6
.
sin
Squaring and addini

cos
2
Y
r.f
cos
cos
/)
2
;
sin
2
<z>
cos1 Cr-xsi;
2/3
cos
or
if
sin- (y
we transform from polar

o
to rectangular coordinates
o
c
cos*
+ -r-r^y = ^ (0
This equation and the equation of the sphere

A9
+.!*
+ **
nr
r2
(2)
determine the spherical

cylinder and the sphere.
ellipse, as the intersection
of a right elliptic

1
107
68.
To find
the equations to the helix.
Whilst the rectangle revolves uniformly about so that the parallel side generates the surface of a right circular cylinder, the point P moves uniformly along CM, and generates a curve called a helix.
Definition.
side
its
AB,
ABCM CM
Let
AB
be the axis of
xz
let
P and
both be at
=n
Also
times the velocity

.-.
and when the rectangle is in the plane on the axis of x, and let the velocity of of M.
z,
PM =
y,
;/.
arc.
DM.
and
let
AN =
jv,
NM
1
PM =
2
j
z be the coordinates of P,
AM =
a the radius of the circular base of the cylinder in the plane xy.
z
na
cos"
and
x^
<r
(i)
are the required equations of the helix. Or we may represent the curve by the two equations
na
forms
cos"
na
sin"
a
or the
(2):
same
in the
x
z
a cos
na
/
a sin
na
n)
z
and
z n-\
since cos
na
the
cos
2m
7t
-\
z \ and na J
sin
/
sin
2m
na
na
same values of x and y correspond to an infinite number of z. The equations (i) (2) and (3) show that the projec tions of the helix on the planes xz, and yz give the curve of sines, and the projection on xy is the circle.
values of
PLATE T
PLATE
II
7 DAY USE RETURN TO DESK FROM WHICH BORROWED ASTRONOMY, &ATHEMATICSSTATiSTICS LIBRARY
This publication is due on the LAST and HOUR stamped below.
1
,
DATE
RB
17-60m-3, 66
(G1105slO)4188
General Library University of California Berkeley
BERKELEY LIBRARIES
814023
y -
<#
UNIVERSITY OF CALIFORNIA LIBRARY

Elements of Solid Geometry-1891

Caricato da

Informazioni sul documento

Descrizione originale:

Copyright

Formati disponibili

Condividi questo documento

Condividi o incorpora il documento

Opzioni di condivisione

Hai trovato utile questo documento?

Questo contenuto è inappropriato?

Copyright:

Formati disponibili

Elements of Solid Geometry-1891

Caricato da

Copyright:

Formati disponibili

Irving Strin^han

PROFESSOR OF MATHEMATICS, UNIVE-RSITY OF VIRGINIA.

UNIVERSITY PUBLISHING COMPANY

NOTES ON SOLID GEOMETRY.

and O2 the plane xOz

PM, PN, PR from

the zOy, zOx,

lines parallel to the axes CXv, Oy system of coordinates in space is called

are the edges,

and the point P

the opposite corner, so that

three planes xQy,

x y, and z and yOz are called the xOz,

* For Figures see Plates

NOTES ON SOLID GEOMETRY.

in front of the plane

the eight different angles thus

In the sixth Octant, P 6 is In the seventh Octant, I\ is In the eighth Octant, P8 is

and the lengths of

Construct the points Construct the points

state the contrary.

other properties of orthogonal projections

NOTES ON SOLID GEOMETRY.

projection of a line on a plane projections of its points on the plane.

the line containing the

line or several lines connected together enclose a plane

area, the area enclosed

by the projection of the

limited line on the second line.

right line on any plane

multiplied by the cosine

the plane of pro is the projec PQ with the plane

R, then by PQ with MN.

Q draw QR parallel and the angle PQR

begin with a triangle of which one side

to the plane of projection.

x AD, and the

area of the projection

NOTES ON SOLID GEOMETRY.

of the triangle draw

plane of projection meeting AB in D. between the planes, then from i C

= BCD cos = ABC cos

A D = ACD cos 6 A B D -B C D =(ABD - BCD) cos

upon another right

to the first line multiplied

the angle between the

the given line and the perpendiculars

with Ox, and

be three points P, P the projections of

joined by the right lines

according as the cosine of the angle above

lygonal line on a straight line of the other sides on that line.

USEFUL PARTICULAR CASE.

rectangular coordinates are (x,y, z] and (x

one of the points P

be at the origin then

FIND THE RELATIONS BETWEEN THE COSINES OF THE ANGLES

WHICH A STRAIGHT LINE MAKES WITH THREE RECTANGULAR

= r cos a y = r cos ft r = r (cos a -f cos ft

determine the direction of the line in rectan

Only two of the angles

can be assumed at pleas

ure, for the third, y, will be given