ON POLYZOMAL CURVES.
557
4 14]
see ante, Nos. 87 et seq.—it there appears that (unless the centres A, B, C, D are in
a line) the condition signifies that the four circles have a common orthotomic circle;
and when we have also
l m n p „
- + t- + - + 3 = 0.
abed
The formulae for the decomposition are given ante, Nos. 42 et seq. Writing therein
21°, 23°, (5°, iD° in place of U, V, W, T respectively, it thereby appears that the tetra
zomal curve Vi2l° + Vm23° + \/n(i 0 + = 0, breaks up into the two trizomal curves
where
+ Vm$° + ^1° = 0, V42l° + VwU3° + Vw 2 (S 0 = 0,
, a P_
+ d Vi’
and where we have
A
1 no
>
II
a p
+ d Vi’
A ='
Vwj
= V m
“ \!bed
£b
1
A,
V nio = '
\A
= A
+ / \/bed
p c
1 c
fm,
A 2 = '
have
L m,
n,
= 0,
L m 2
—f TT "•
a b c
a + b
'm,
Article Nos. 198 to 203. Gases of the Decomposable Curve, Centres not in a line.
198. I assume, in the first instance, that the centres of the circles are not in a
line; we have the following cases:
I. No further relation between l, m, n, p; the order of the tetrazomal is = 8;
the order of each of the trizomals is = 4, that is each of them is a bicircular quartic.
II. v7 + Vra + Wn 4- fp = 0 ; the order of the tetrazomal is =7, that of one of
the trizomals must be = 3.
To verify this, observe that we have
A 4- Vwj 4- = A 4- Vra + fn 4- (c fm — b Vn),
or substituting for \!l 4- Vra + A the value — fp, this is
Vi
= |a fp — d A 4- sj^(cVm-b A)j,
and similarly for V/ 2 + Vm 2 4- Vn 2 , the only change being in the sign of the radical
/ad
V be
. But from the two conditions satisfied by l, m, n, p it is easy to deduce
(a fp — d Vif —(c Vm — b Vnf = 0,