. DEVICE XGP    
.spacing 20*5 mills;
.EVERY HEADING(,{PAGE},)
.AREA TEXT LINES 4 TO 40
.FONT 1 "MBIG[HHA,LCS]"
.font 2 "BMETI[HHA,LCS]"
.font 4 "BMUS[HHA,LCS]"
.font 6 "MBIG[HHA,LCS]"
.!XGPLFTMAR←400
.PORTION MAIN;
.PLACE TEXT;
.PAGE FRAME 138 WIDE 106 HIGH;
.COMPACT
.COUNT PAGE FROM 1 TO 999;
.<< THICKEN UNDERLINE >>
.AT "↓_" ⊂ SNEAK(BEWARE("'177'1'46"))⊃;
.AT "_↓" ⊂ SNEAK(BEWARE("'177'1'51'4'4"))⊃;
.<< Put in a footnote. >>
.
.COUNT FOOTNOTE INLINE FROM 1 TO 999 IN PAGE PRINTING ⊂"*****"[1 TO FOOTNOTE]⊃
.<<	(IF THISDEVICE = "XGP" THEN "%51%*" ELSE "[1]");>>
.
.FOOTSEP ← "__________";
.AT "$$" ENTRY "$"
.	⊂
.	NEXT FOOTNOTE;
.	FOOTNOTE!;
.	SEND FOOT
.		⊂
.		BEGIN "NEXT FOOTNOTE"
.		SELECT 1;
.		SINGLE SPACE
.		SPACING 0 MILLS
.		INDENT 0,0,0;
.		(FOOTNOTE! & " ");
ENTRY
.		END "NEXT FOOTNOTE";
.		⊃;
.	⊃;
.
.TURN ON "%↓_↑↓[&]","α"
.at "!!" txt ";"	⊂
.("↑[%3"&"txt"[1]&"]&↓["&"txt"[2]&"]%*");
.COMMENT ("txt"[1]&"↑[%3"&"txt"[2]&"]&↓["&"txt"[3]&"]%*");
.  ⊃
.PAGE←122
.NEXT PAGE
.FILL INDENT 12
	Because of the slight changes in the presentation of the motive
in bars 65-69, we now find somewhat different harmonic details.  The
harmony of bar 65 is first heard as ↓_c_↓: I, the disjunctly approached A%4F%1
at the end changing the function to VI%47%1.  Continuing on this basis,
bar 66 will give ↓_c_↓: II acting at same time as ↓_f_↓: %4N%1VI, which then becomes
↓_f_↓: IV%47%1 at the end of the bar.  (Note that a B%4F%17 chord rarely can be
truly functional in ↓_c_↓.)  Bar 67 then has ↓_f_↓: %4N%1VII to V%47%1, which works
equally well as ↓_g_↓: %4N%1VI to IV%47%1. This leads 
to ↓_g_↓: %4S%1VII%47%1 to V%47%1 in 68 and
then I at the beginning of bar 69.
.BEGIN VERBATIM

Figure 93g
.END
.CENTER
%6⊂⊗⊃L[α%0.04,α%-2.06](17,22):N93FG.PLT[C9,LCS]⊂⊗⊃%1
. << SIZE .9 >>
.SKIP 4
.FILL INDENT 12
	However, it hardly seems necessary to show these changes of
%4N%1VI to IV%47%1, etc. in bars 66-68, the root of each seventh chord being
delayed until the last eighth note.  But if the same reasoning is
applied to the slightly different structures of bars 65 and 69, what
is first heard clearly as a minor I chord then is "completed" 
as IV%47%1 of a major key.
Thus a momentary tonic of ↓_E%4F%1_↓ is implied, overlapping with the ↓_c_↓
of bar 65 and the ↓_f_↓ of bars 66-67.
.BEGIN VERBATIM

Figure 93h
.END
.CENTER
%6⊂⊗⊃L[α%0.04,α%-5.66](17,22):N93FH.PLT[C9,LCS]⊂⊗⊃%1
. << SIZE .9 >>