250 	GO SUB 510
	260 	LET Arcs=Arcs-1: IF Arcs=0 THEN STOP
	270 	LET MM1=M1
	280 	LET M1=M1*M3-M2*M4
	290 	LET M2=MM1*M4+M2*M3
	300 	GO TO 210

	500 	REM 'DRAW A LINE' from last position to X,Y
	510 	LET PLOTx=PEEK 23677: LET PLOTy=PEEK 23678
	520 	LET dx=SGN X: LET dy=SGN Y
	530 	LET X=ABS X: LET Y=ABS Y
	540 	IF X> =Y THEN GO TO 580
	550 	LET L=X: LET B=Y
	560 	LET ddx=0: LET ddy=dy
	570 	GO TO 610
	580 	IF X+Y=0 THEN STOP
	590 	LET L=Y: LET B=X
	600 	LET ddx=dx: LET ddy=0
	610 	LET H=B
	620 	LET i=INT (B/2)
	630 	FOR N=B TO 1 STEP -1
	640 	LET i=i+L
	650 	IF i < H THEN GO TO 690
	660 	LET i=i-H
	670 	LET ix=dx: LET iy=dy
	680 	GO TO 700
	690 	LET ix=ddx: LET iy=ddy
	700 	LET PLOTy=PLOTy+iy
	710 	IF PLOTy <0 OR PLOTy > 175 THEN STOP
	720 	LET PLOTx=PLOTx+ix
	730 	IF PLOTx <0 OR PLOTx > 255 THEN STOP
	740 	PLOT PLOTx,PLOTy
	750 	NEXT N
	760 	RETURN



NOTE ON SMALL INTEGERS AND -65536.

1. Small integers n are those for which -65535 is less than or equal to n which is less
than or equal to 65535. The form in which they are held is described in 'STACK-BC'.
Note that the manual is inaccurate when it says that the third and fourth bytes hold n
plus 131072 if n is negative. Since the range of n is then -1 to -65535, the two bytes can
only hold n plus 131072 if it is taken mod 65536; i.e. they hold n plus 65536. The
manual is fudging the issue. The fact is that this is not a true twos complement form (as
the form n plus 131072, in other circumstances, could be). Here the same number can
stand for two different numbers according to the sign byte: e.g. 00 01 stands for 1 if
the sign byte is 00 and for -65535 if the sign byte is FF; similarly FF FF stands for
65535 if the sign byte is 00 and for -1 if the sign byte is FF.

2. Accepting that negative numbers are given a special 'twos complement' form, the
main feature about this method of holding numbers is that they are ready for 'short
addition' without any further twos complementing. They are simply fetched and stored
direct by the addition subroutine. But for multiplication they need to be fetched by
INT-FETCH and stored afterwards by INT-STORE. These subroutines twos com-
plement the number when fetching or storing it. The calls to INT-STORE are from
'multiply' (after 'short multiplication'), from 'truncate' (after forming a 'small integer'
between -65535 and 65535 inclusive), from 'negate'/'abs' for the 'integer case' and from
'sgn' to store 1 or -1. The calls to INT-FETCH are from PRINT-FP to fetch the integer
part of the number when it is 'small', from 'multiply' twice to fetch two 'small integers',
from 'RE-STACK' to fetch a 'small integer' for re-stacking, from 'negate'/'abs' to fetch
a 'small integer' for manipulation and from FP-TO-BC to fetch the integer for transfer
to BC.