L.C.F.B..
14 JANi98L ImperialTobace' Li ited/Limi((~-,
VIC
FILE jqd,~-
PASS ................
COPY . ......................
~v ? co/,
. ................. ow (514)932-611;1
....................... W
K1 ... orp-A. P0. 11:4: 31m,
....................... ewe C.04" Gibl~ P,
e
January 7, 1981 RESTRICTED
Dr. G.A. Few
British-American Tobacco Co. Ltd.
38 Regent's Park Road
Millbrook
Southampton S09 1PE
England
Dear George,
Following our telephone conversaiton of December
19. 1 am sending you under
separate cover a further carton of each of the
brands, Accord Ultra Mild K.S.
(B & H) and Matinde Extra Mild K.S. (ITL). I am
taking the liberty of
including with them a carton of a recently
modified 8 & H brand, Viscount I
Ultra Mild K.S., for reasons which will be
clarified later in this letter.
Over the past 9 months, Benson & Hedges have been
extraordinarily active in
the very low tar market.
Their new or modified brands are:
Package Tar (mg)
Accord Ultra Mild K.S. 3
Accord Ultra Mild Menthol K.S. 3
Viscount I Ultra Mild K.S. .0.7
Viscount I' Ultra Mild Regular 0.4
Viscount Extra Mild K.S.* 3
Viscount Extra Mild Menthol K.S. 3
Viscount Extra Mild Regular 5
Belmont Ultra Light K.S. 3
40 2
C r-,
1,S pe.14&0
d) C., "07-9 Ic-lo e cc"
C2-
110
co
co
BATCo document for Province of BritiSh Columbia 10
November 1999


- 2 -
All brands except that marked with an asterisk
contain the light yellow
puffed tobacco (generally 11-12%, but the blend
for Viscount I Ultra Mild
Regular looks to contain a considerably higher
level). Puffed tobacco had
been found previously, but not consistantly, in
earlier versions (5 mg tar)
of Visqount K.S. and Regular.
We have the data from the original brand breakdown
of most of these brands
(given foryourinterest in the attached table), and
we are in process of
getting a more complete picture with new samples
from the market.
Apart from the normal physical and chemical
analyses, we are undertaking the'
following tests on some or all of these brands:
1. We will determine the level of puffed tobacco
in each blend, and will
isolate, by flotation on hexane, enough of the
puffed tobaccos to
compare % reducing sugars, nicotine and nitrate
(suggested by your
work). We will also compare the puffed tobaccos
for visual similarities.
Our suspicions are that B & H may be having short
strips expanded instead
of cut lamina, from the "crushed cornflake" shape
of the particles. As
I mentioned on the phone, we will be looking at
the degree of commonality
in chemistry of the puffed tobaccos. If this
proves consistant, it
should rule out a contribution of the puffed
tobacco to the unique
taste character of the smoke from Accord, since
none of the other
brands have the Accord taste in our opinion.
2. We will determine the nitrate content of the
total blend for four of
the brands (compared to those of Matinde Extra
Mild and Medallion).
This is again picking up on your finding of very
high nitrate content
of the unexpanded portion of the Accord blend.
The brands we will test for this are:
Accord Ultra Mild K.S.
Viscount I Ultra Mild K.S.
Viscount I Ultra Mild Regular
Viscount Extra Mild K.S.
Visually the blends from Accord and Viscount I
Ultra Mild K.S. look
identical, and they have the same reducing sugar
and nicotine contents
(12% and 2.0% respectively). The blend of Viscount
I Ultra Mild Regular
is a much lighter colour, also with more puffed
tobacco. The blend of
Viscount Extra Mild K.S., which we believe is
unaltered from the previous
version, has no puffed tobacco but is dark in
colour. It has a similar
reoucing sugar level to Accord (12%) but much
lower nicotine content (1.3%).
3
CXD
BATCo document for Province of British Columbia 10
November 1999


- 3 -
3. We will compare the smoke pH and % extractable
nicotine of the smokes
from the brands discussed in 2 above, to see if
these reflect the
suspected air-cured character of some brands. -
4. We will endeavour to Intercompare the headspace
or filter-tip volati.les
of four brands by the GR & DC technique. The
brands are:
Accord Ultra Mild K.S.
Viscount I Ultra Mild K.S.
Viscount I Ultra Mild Regular
Matin6e Extra Mild K.S.
The reasons for this choice stems from their
subjective smoking
character. I have previously described Herb
Roubicek's comparison of
Accord and Matinde Extra Mild. We now feel that
Accord's taste has
some air-cured character but with unusual
'high-notes' which might
derive from some added flavour or non-'Canadian
flue-cured' enhancer.
In comparing the smoke from the tobacco sections
of Accord and Viscount I
Ultra Mild K.S., we find that the smoke from the
latter also has a
similar air-cured character but without Accord's
unusual 'high-notes'.
Thus we feel this Viscount version may be a good
'control' for the
headspace comparison. This is why I feel it will
be valuable for you
to have some of these cigarettes, for the studies
you have offered to
arrange.
We will also look at Viscount I Ultra Mild
Regular, because its tobacco
section gives an unusual smoke taste character -
not air-cured but with
some flavoured notes.
As I mentioned on the phone, Accord is the brand
which is creating 'waves'
in our company, because of the signs of inroads it
is making in market
share in this tar segment, showing positive gains
from all competing
brands, including Matinde Extra Mild. A blind
consumer test is being
prepared of Accord versus Matinge Extra Mild to
get a better appreciation
of perceived differi-n-ces.
We would very much appreciate any further
contribution that GR & DC can
make relative to Accord's unusual taste character.
As I mentioned on
the phone, I feel that comparison of the
headspaces of the brands by
John Green's method should be valuable, especially
knowing GR & DC's
potential for MS identification of any unusual
peaks. Another suggestion
discussed on the phone was the comparison of O/WP
scans of the smokes
using FID and nitrogen specific detectors, to look
for possible enhancement
of pyridines/pyrazines.
4
CO
00
QPJ
BATCo document for Province of British Columbia 10
November 1999


- 4 -
I look forward to any further results, comments or
suggestions on these
products, please don't hesitate to call me if you
need-more information.
For example, you suggested reevaluation of your
blend analysis data would
be valuable.
With best wishes,
Yours sincerely,
T.A. Smith
TAS/dn
Encl.
.c.c. Pr. L.C.F. Blackman
CXD
CO
BATCo document for Province of British Columbia 10
November 1999