APR 1 9 1965

Crops Division

Fort Detrick
Frederick, Md. 21701
'65 April 15

Dear Al (and other notationers):

By this time I hope you have had a chance to relax and exchange some
topological gems with Dr. Lederberg.

You may recognize one of the enclosed sheets as a direct copy of a
tantalizing "chemicotopological hiatus" (a vacancy in our old analyses) that
you posed in 1963. I hope you will give my enclosed tabulation the same bit
of attention that we gave your ideas thens in the listed "parenthetical"
notations, your added specifications are separated from the old ring marks
‘ by the slash, for this pedagogial point: which instructions are easiest?

Howard Bonnett's concern also is my apprehension: we are not topo-
logical diagnosticians, and if we cannot deduce correct ring loops from a
diagram that reveals all symmetry relations, we are not going to do one bit
better by redrawing different kinds of diaprams.

The lookup table (my answer to this hiatus) is well along the way.
Just a few hours after I had free time with the BioMath machines, I had an
arithmetically closed deck of the first 1330 tricyclic combinations, and by
fast sorting had excluded invalid (polynuclear, etc.) combinations.

I really was disturbed by the failure of any of you to see these
absolutely infallible simple circuit definitions as I always have seen them.
The nonconsecutive line segments are a sharply defined set that begins thus:

AC, AD BD, AE BE CE, AF BF CF DF, AG BG CG DG EG, AH BH CH DH EH FH, etc. ee
1, 2 3, h S 6, 7 8 910, 11 12 13 14 15, 16 17 18 19 20 21, etc. oe

so it is child's play to make checklists from these letter pairs alone.

When you complain that the "locant links" do not show ring sizes,
please remind yourselves that these "smallest sets of smallest rings" are
pure figments of chemist's minds-- hence trouble. And if we decide first
on smallest sums of ring numerals, then on smallest other things, where is
there any contradiction such as you implied in your note of August 22, 1963:
"The SMALLEST AND FEWEST rings turns out to be impossible~- sometimes you
can't have BOTHJ'' My pathetracers give ALL possible alternatives of longest
chains, and alternatives are eliminated by specified minimum measures, no?
The only “arbitrary assertions" are these chemically traditional choices,
with topological exactness: smallest number of rings, smallest R.N. sun.

If you study the enclosed '65Ap15 tabulation carefully, you will see
absolute relations, such as this: Your "enclosed" (underlined) locants are
(1) not in the linking recitation if they are intermediate-chain points, or
(2) recited only once if they are chain terminals (a or k in this example ).
Spiro/quadrivalent points also are revealed directly in these recitations.

Am I getting through to anyone?
Sincerely yours, pill
{
SAMPLE TOPOLOGICAL ANALYSIS OF 8=jPOINTED HEXABRANCHED PENTAGON:
(Pictured as "8=/POINTED TRIANGULAR PRISM" on drawing sheet)
(Forms trigonal prism when bivalent points are removed )

PATH entity /~

 

 

NONCONSECUTIVE CORRESPONDING LINE- SUM OF SUM OF SUMMARY OF RING costes, FINAL
LOCANT LINKS FORMULA NOTATION FUS.L, RING # CIRCULT-CHECK BY POINT-TYPES & NO. \DRG #; RANK
1 AC AF BH EH (355 BL/G2AF H) 2.382 17 = azbgceadegfzgoho = 25364" = 17 17
2 AC AF DG BH (35 DyS/E2aF H) 7 17 agbgegdzeafzggh —- 121,635 ly 10 06h
3 AC BF AH EH (3. B55 E4/G2BF H) 9 17 agb3egdepfzezh2 «1253.64 " 2 2200 oh
h, AC DG BF AH (3 Dh BSS/E2BF H) 8 17 agb3egdgeaf3goh 211,635 8 18 20
5 AD AG CH FH (uS5 F3/B2AG H) 9 17 agbgeadgefpe3hp «446435211 -1 2522
6 AD BF EG AH (uS5 E35/H2AG H) 8 17 a3bgedgeofogzho 34641125 | 10 19 18
7 AD BF EG CH (LS E3 B5/A2BF H) 9 17 apbscadzert 3e2h 64341125 ~1y 23 23
8 AD CG AH FH (4 C55 F3/B2CG H) 11 17 agbge3dgefog3h2 = 46135211" -2 28 28
9 AE AF BH DH (535 Bu/C2AB H) 5 17 azb3epdge9f eh, 14641253 7 2 2
10 AE AF CG BH (535 BL/3ABC H) 5 17 ajb3e34epf 96h 14352146 6 1 1
lL AE AF CG DH (535 Ch/B2AC H) 6 17 agbge3dgefoggh = - 25311146 9 5 5
12 AE AG CH DH (545 03/B2AC H) 6 17 agbge3dgepfgoho «3521641 -7 6 6
13 AE BF CH DH (Su. BS C3/A2BC H) 7 17 — agb3c3dgeqfoehp 641,352 -9 13 10
Uy AE CF AH BH (5-Ch53/D2AE H) 6 17 agbgeadge3foghy 146,352 3 7 7
15 AE DF AH CH (5 D354/B2AE H) 7 17 agbgendze3fogh, —*464411253 -11 1 8612
(ete., to 30) .
All possible notations are ranked by: (1) ring sum, (2) locant sum, and (3) locant-citing order. 165Mar30/WIW

The cited ring & locant is the SMALLEST & LOWEST going through each newly formed link.
Note that the "point-type" paths occasionally are ambiguous, unlike the "Locant links" and generated notations.
jot “what the eye can see.

%
3

oe

ao
-¥

c.

wv

Ory wh» feel they CON succes ely encede

22, \963

Avg.
EGS To :

es PUb 24 KECI

-2- 63-7

5) I hope this new analysis may help convince any doubters (hiya, Bill) that
my analytic approach to structural fomulas via the concept of bridging —
atoms and multicyclic points is a sound one, for purposes of this hotation.

6) I believe this analytic approach could be applied to other codes or notations
since it is really independent of the particular notation we're using in most
ways. Does anybody know how Dyson (opps, I mean the IUPAC notation) handles
this problem of WHICH structural formula to encode in compounds like these?

ions I hope you'll all take the time and trouble to plow through this material.
I'm sure there are still loose ends, 80 please let me know what further difficulties

you turn up.

I'l) be in Columbus, Ohio from Sunday Aug. 25 through Wed. the 28th doing ay
bit to present the final report of the NRC project to the parent committee. I
know I'11 see some of youthere. Whether or not I get to go on to the Bureau of
Standard's (Tauber's) seminar the next week depends on whether or not Mrs. Brownson
and Karl Heumann say it's okay for me to go on there from Columbus (at my own
expense) and still charge my return trip ticket from Columbus te San Francisce
to the NRC project. I just don't know what the regulations are on this sort of
thing. If I do get to stay over, could you and I get together, Bill? Over the
Labor Day week end, maybe? Or sometime? fhege seems to be quite a lot we need
te talk over in order to get the manual revision going. \

”

 

 

 

1) sacs ga SUSe (196) Sincerely,
OA
. Elbert G, Smith
= ae SUGGETTED “ANAUITICAL
, ( APROACH" -
—> (2 > re

 

 

3) Ring Tradex OT!

ep ~
‘D-

 

 

TT,

WIN: FIRST IDENTIFY KinDS oF PoINT-CoNNECTIONS 4
Ape. I 965

 

 

 
 

 

“oO an nn Ww Fw nN e

18

All possible notations are ranked by: (1) ring sum, (2) locant sum, and (3) locant-citing order.

(> FrexenT)

SAMPLE TOPOLOGICAL ANALYSIS OF CONNECTIVITY TABLES FOR A ,HEXABRANCHED PENTAGON: .

 

PATH IDENTITY Sv” 1.

 
 

25521736UK6! = 2 or 15

"

h

3
8

1

5

6

NONCONSECUTIVE CORRESPONDING LINE~ SUM OF SUM OF SUMMARY OF RING
LOCANT LINKS FORMULA NOTATION FUS.L. RING # CIRCUIT-CHECK BY POINT-TYPES

AE AI DJ GK (565 G5/H3AEI K) 12 21 = a3;bedze3fgoh2izjok

AE AI FJ CK (S656/\AREF K) h 22 = albocgde3foghi3zjok 1,636)253217

AE BH CK GK ($6 B6 BS/ALBBCH K) 6 22 = agbjczdeofgohzijlky 64463712552" =15
AE CH BK GK (5 C6 B6 BS/DUBCCH K) 8 22 = absedoeteshgi dke 64463712552" = Uy
AE DI AJ GK (5 D65 G5/H3DEI K) 13 21 = agbedye3fephyis.ok 255217 364L6

AE DI FJ BK (5 D6 DS6/AMDEEF K) 10 22 = apbgedye),fyghigjgk 63641255217

AF AI Ed CK (655 C5/D3AEF K) 6 21 = agbegdze3f3ghigjgk 1736K2552h6

AF BH DI CK (65 BS C5/A3BCD K) 7 = apbze3d3efzehgtp dk 64217364255 = 10
AF DH BI CK (6 DS BS BS/E3BCD K) 9 21 = abzczdzecfzehaiajk 7126364255 = 16
AF EI Ad CK (6 E555/B3AEF K) 8 al = agbpopde3f ghip Joke 46371255216

BF AI CK GK — ( B566 B5/ALBBCF K) 6 22 = agbjc3defzeahinjky 712552h636) = 18
CBF CH Ad GK ( BS B56 G5/I3BCH K) 12 21 = abgezdefogghzinjak 7125521636 = 17
BG AI CK FK ( B656 C5/ALBBCG K) 7 22 = agbje3defogghinjkg 64255246371 = 12
BG DH AJ CK B6 DS65/A E F2BG K) 8 22 = agbsendzenfze hol Jkblh25521736 = 13
CG AI BK FK ( C565 BS/A3BCG K) 7 21 = agb3zc3zdefogshigjky 64255217364 = 11
CH AI DJ FK ( 665 CS Do/e3CDH K) 1 21 = abezdzepfoahzigjgk 552h63712h6 = 7
DH AL CJ FK ( D565 C6/A BE2CD KK) 9 22 = apbge3dzenfpghaipSgk552h636L217 =
DH AI EJ BK ( D56 DS5S/A3DHI K) 10 21 = agbpedyeofeh3zak 637125526

(!

9

-12

FINAL |

15
10
18

w

17
6

‘C5 Ap.1S/ WW