_
3B 275
.G23
Sopy 1

w..S. DEPARTMENT OF AGRICULTURE.
BUREAU OF PLANT INDUSTRY—BULLETIN NO. 100, PART IV.

B. T. GALLOWAY, Chief of Bureau.

ETHODS OF TESTING THE BURAING
QUALITY OF CIGAR TOBACCO,

WIGHTMAN W. GARNER,

ScreENTIFIC ASSISTANT, PLANT BREEDING INVESTIGATIONS.

IssuED JUNE 9, 1906.

WASHINGTON:
GOVERNMENT PRINTING OFFICE.
—vY

1906. re
aA

Honograph

CONE NaS:

Htrastangco GUC a1 eae re ere me ee eps One oe Slate Jay ai Sig Siete Siok seer ew
"Tees Se TNO NLS a ES WEY ee ss Pa ee eae aca at te eae ae eee r= an aA Pe
The effects of the filler, the binder, and the wrapper on the burn of the cigar-
Testing the capacity for holding fire and the evenness of the burn
Testing the burn of cigar-filler tobacco

LAOS PRON Seem

PLATES.

Prare I. Variation in burn of wrappers due to different fillers...-..-. = z
II. Variation in burn of fillers due to different wrappers and binders
‘ » vt Z 4

TEXT FIGURES. ee

4 =

_ METHODS OF TESTING THE -BURNING
QUALITY OF CIGAR TOBACCO.

INTRODUCTION.

As has been pointed out in previous publications of the Bureau of
Plant Industry, a systematic effort is being made to improve the qual-
ity and yield of the tobacco crop by employing the latest and most
approved methods of selection in the old varieties and by creating
and establishing new strains possessing to a marked degree those
characteristics most to be desired in the various classes of tobacco
which the market demands. This work necessitates the careful test-
ing of a large number of types, as well as many individual selections
from each of these types, and for this reason it is very desirable to
have at our command methods capable of showing with certainty even
slight differences in the essential qualities of the various samples to be
examined. It is our purpose to make a careful study of the subject
of testing tobacco from a practical standpoint, as well as the relation
of the chemical composition of the leaf to its good and bad qualities.

In judging the merits of a cigar tobacco, due regard must be had
for the particular use for which it is intended, since the finished cigar
consists of three distinct components—the filler, the binder, and the
wrapper—each of which must possess certain characteristics. The
“In the tobacco breeding experiments conducted by the Plant Breeding Inyestiga-
tions of the Bureau of Plant Industry, particular attention is being given to the
improvement of cigar tobaccos, including specially high-grade wrapper and filler
types. In connection with these experiments, which are being conducted by Messrs.
A. D. Shamel and W. W. Cobey of this Office, it has been found necessary to com-
pare the characters of a large number of selected individual plants to determine
which ones are superior in their important characters. The means and methods
heretofore used in making such comparative tests were very imperfect, and one
important preliminary part of the work is to devise special pieces of apparatus which
will enable accurate tests to be made. The devices described by Dr. Garner in the
present paper it is believed will greatly facilitate such testing and add to the accuracy
of the results. The preliminary notes given by Dr. Garner on the influence of
wrapper, binder, and filler on the ‘‘burn”’ of cigars open up an important field of
investigation in connection with the testing and breeding of different types of
tobaeco.—Hersert J. Wesper, Physiologist in Charge of Plant Breeding Investigations.

5

6 TESTING THE BURNING QUALITY OF CIGAR TOBACCO,

filler must have, above all else, a fine flavor and aroma and a good
‘“burn.” In the case of the wrapper leaf there are a number of
requirements to be met, among which are sufficient elasticity, proper
color, size, and shape, fineness of veins, freedom from objectionable
flavor and taste, a fime ‘‘erain,” and a good burn. Many of these
qualities can be determined by simple inspection, without the use of
any specific tests, while others require special laboratory methods.
The present article has to do only with the practical methods of testing
the burn, deferring to a later day a consideration of the chemical
characteristics of the tobaccos which have been tested.

There are several elements which go to make a good or bad burn,
chief of which are the capacity for holding fire, the evenness of the
burn, the color of the ash and its firmness, the coaling or carboniza-
tion, and the ‘‘ puckering” of the leaf immediately in advance of the
burning zone of the cigar. The final test of any cigar tobacco must,
of course, rest in the smoking of the manufactured cigar, but, while
this gives a direct means of determining the character of the ash, it
does not furnish accurate information as to the evenness of the burn
or the fire-holding capacity of any one of the components of the cigar
except with reference to the other two particular components used in
the experiment. This is particularly true of the wrapper, as was
shown by special experiments carried out to observe the effect of
using different fillers and binders with the same wrapper. The result
of these experiments will be more fully discussed below. Again, it
should be remembered, in this connection, that cigars made by the
same workman and from the same lot of tobacco often vary widely in
their burn owing to the impossibility: of avoiding unevenness in the
filler, and this source of error can only be eliminated by several times
repeating the experiment. It is evident, therefore, that, in order to
get reliable data concerning the relative merits of different wrappers
with respect to their burning qualities, the cigar test must be supple-
mented by some other method capable of giving sharp distinctions as to
the fire-holding capacity and evenness of burn.

A method which has long been in use is to ignite the leaf by means
of a lighted cigar or a slow-burning match deviced by Nessler,“ and
note the number of seconds during which it continues to glow. The
mean of several tests is taken as a measure of the capacity for holding
fire; but the variation in the results obtained, even upon a single leaf,
is so great that little reliance can be placed upon the figures except in
avery general way. In this method no account is taken of the area
of the leaf burned, and the wide differences obtained on a single leaf
are due principally to the fact that frequently the ignited zone soon
ceases to glow except for one or more very small streamers, which

4 Landw. Vers. Stat., XI, 399.

THE SMOKING TEST. fe

continue to burn for a much longer period, thereby giving results
altogether out of proportion to the true burning qualities of the
tobacco in question. Another serious objection to the method is found
in the interference of the veins of the leaf; for it seldom happens
that the glow can cross these veins except around the outer edges of
the leaf, while in the cigar the veins always run longitudinally and
so do not interfere with the burn.

The rational procedure would seem to be to test the burn of the
leaf when wrapped in some such form as is actually found on the cigar,
but without the use of binder or filler. We have devised a method of
this kind in which the leaf, after being properly wrapped and dried, is
burned with the aid of a slow current of air. The current of air com-
pensates in a measure for the absence of the filler and binder, while its
use obviates the unavoidable irregularities of the latter. A detailed
description of the apparatus used and the method of carrying out the
operation will be found on page 11. This test, combined with the
smoking of the cigar, has enabled us to accurately classify a large
number of samples of wrapper leaf with respect to their burning
qualities.

THE SMOKING TEST.

It is evident that no two persons would smoke a cigar in exactly the
same way, nor would the same individual smoke two cigars under
exactly similar conditions. It is necessary, therefore, to use some
means of smoking the cigars artificially in order to eliminate the per-
sonal equation and secure uniformity of conditions. Dr. E. H. Jen-
kins, in the Annual Report of the Connecticut Agricultural Experi-
ment Station for 1892, has described an apparatus for smoking cigars
which was devised by Mr. 8. L. Penfield, of Yale University. The
** pull” on the cigar is secured by means of an aspirator which is filled
by a continuous inflow of water and emptied at regular intervals by :
siphon. We have modified this apparatus in a number of details in
order to adapt it to our needs, and we give herewith a description of
the form and dimensions which we have finally adopted for use in
our investigations. In this apparatus as many as four cigars may be
smoked simultaneously, while held in such a position that they may
be readily compared throughout the operation.

By reference to the accompanying illustration (fig. 1) it will be seen
that the holders (a, }, c, and @) for the cigars are so arranged that
they all lie in the same vertical plane, each one 2 inches above and
having its horizontal arm 2 inches shorter than the next lower. A
screen with a white covering is placed immediately in the rear of the
holders to serve as a background, thereby facilitating observation of
the character of the ash. Between the flask bearing the holders and
the aspirator and connected with these by means of glass tubing is a

8 TESTING THE BURNING QUALITY OF CIGAR TOBACCO.

check valve consisting of a T-tube (2), the lower arm of which dips
beneath the surface of the water ina suitable vessel (7), thus prevent-
ing a backward draft through the cigars while the aspirator is filling.

The smoke which escapes through this valve has a very cienenes tee
odor: hence, it is well to use a bottle fitted with a stopper with two
holes, through one of which passes the lower arm of the T-tube (9),
while through the other passes one end of a long tube (4), which car-

ries away the obnoxious fumes. The aspirator consists of a glass cyl-
inder (/), the upper end of which is fitted with a rubber stopper

Fig. 1.—Apparatus for testing the burning quality of cigars: a, }, ¢, d, holders for cigars; e, T-tube,
the lower arm of which, g, dips beneath surface of water inf; h,, tube for leading away tobaeco
smoke; i, aspirator; k, tube leading to water supply; 0, p, long and short arms of siphon; 7, tube
connecting cigar holders with aspirator; m, tube, with bulb, attached to long arm of siphon; 7,
flask carrying the cigar holders.

containing two holes through which pass tubes leading to the water sup-
ply (4) and to the flask carrying the cigar holders (/), respectively.
The lower end of the cylinder is closed with a rubber stopper bearing
the small arm of the siphon (0). The tube furnishing the water sup-
ply is connected with a reservoir provided with a constant-level attach-
ment. To the lower end of the small arm of the siphon is attached a
glass tube (7) bearing a small bulb which materially assists in break-
ing the flow of the water at the moment the aspirator is emptied. The
tubing connecting the parts of the apparatus should not be less than

EFFECTS OF FILLER, BINDER, AND WRAPPER ON BURN. 9

6toSmm. internal diameter; otherwise the tubes will frequently become
clogged by the.condensation products of the smoke. The container
for the cigar holders (7) is filled about two-thirds full with dilute sul-
phuric acid, which serves as an acid wash for the smoke, retaining the
organic bases and thereby further helping to prevent the choking up
of the machine. The lower ends of all the cigar holders should, of
course, extend to exactly the same depth below the surface of the
acid. .

In the machine which we now have in operation, the relation between
the short arm of the siphon and the internal diameter of the aspirator
is such that the volume of water delivered at each emptying of the
latter is 600 ¢.c. This corresponds to an actual capacity of about 450
c.c. for the aspirator, the difference, of course, representing the volume
of water entering the aspirator from the supply pipe while the siphon
isin action. The rate of inflow from the supply tank is approximately
900 c. ¢. per minute. The internal diameter of the long arm of the
siphon is 8 mm., while that of the short arm is 25 mm. The entire
length of the long arm of the siphon exceeds that of the short arm by
40 cm. Anapparatus of the above-mentioned dimensions will smoke
four cigars of the Perfecto type, 4$ inches in length, in about thirty
minutes, a rate which is probably somewhat above that of the average
smoker. The pull on the cigar occurs at intervals of thirty seconds
and continues for a period of ten seconds. The frequency of the pull
is controlled by the rate of inflow of the water from the supply tank,
while its duration is governed principally by the relation between the
diameter of the small arm of the siphon and the volume of the
aspirator.

THE EFFECTS OF THE FILLER, THE BINDER, AND THE WRAPPER
; ON THE BURN OF THE CIGAR.

As preliminary to the use of the cigar test in examining wrapper
leaf, a series of experiments was carried out to determine the relative
effects of the three components of the cigar on the burn. For this
purpose a number of cigars were made by an expert workman, using
four different tvpes of wrapper on each of four different types of
filler. Ina portion of the cigars the binder used was taken from the
same leaf as the wrapper, while in the remainder a sample of Connecti-
cut Broadleaf tobacco was employed for this purpose. These cigars
‘were smoked in the above-described apparatus under conditions as
nearly uniform as could be obtained, and the evenness of the burn and
the character of the ash were carefully noted.

With reference to the evenness of the burn, markedly different
results were obtained when wrappers taken from the same sample were
smoked on different types of filler. A typical case of this kind is shown
in Plate I. The twelve cigars shown were all made from the same

10 TESTING THE BURNING QUALITY OF CIGAR TOBACCG.

sample of wrapper, and in each case the binder was taken from the
same leaf as the wrapper. In the first group (Alive a ne of filler
grown in Texas from Cuban seed was used; in the second group (B), a

heavy filler grown in Ohio from domestic seed; in the third group (CQ),
a filler grown in South Carolina from Cuban seed; in the fourth group
(D), an imported Cuban filler. The wrapper used on these cigars was
a type of Sumatra tobacco grown in Connecticut and had a very good
burn. The Texas and imported Cuban fillers were known to have an
excellent burn, while the South Carolina filler was markedly inferior
in this respect and the Ohio filler intermediate in burning qualities. It
will be seen that this sample of wrapper burned quite evenly when
used with the imported Cuban and Texas fillers, while with the Ohio
and especially the South Carolina fillers the burn was decidedly uneven.

On the other hand, the effect of using different types of wrapper on
the evenness of burn of any one type of filler was less marked (see PI.
II, A, B, and C; also Pl. I, D). The filler used in this experiment was ~
the imported Cuban, while the wrappers were taken from four differ-
ent types of Sumatra tobacco grown in Connecticut. Of these four
types of wrappers, that shown in Plate I, group D, had the best burn
and the one shown in Plate II, group C, the poorest, although little
difference could be seen between the two when smoked on the Cuban
filler. The use of different binders did not cause any marked differ-
ences in the evenness of the burn, as is shown in Plate II, groups C
and D. The cigars used in this experiment were all made from the
same wrapper and the same filler, while in group D a sample of Con-
necticut Broadleaf tobacco was used as the binder, and in group C the
binder was taken from the same leaf as the wrapper.

Another important factor in determining the evenness of the burn is
the proper balancing of the component parts of the cigar. It was
found, for example, that a very light“ wrapper will not give good
results on a heavy filler, even though both of these may in themselves
possess a good burn. It will readily be seen that a very thin wrapper
which burns readily and very rapidly will, when placed on a heavy,
slow-burning filler, tend to burn in advance of the latter, and the effect
will generally be an uneven burn. The same result is obtained when
any cigar is smoked very rapidly, for the reason that the oxygen of
the air has freer access to the outer edges of the burning zone and
under the added stimulus it rarely happens that a cigar will burn
evenly.

As regards the character of the ash, the wrapper and the binder are
relatively of much more significance. It was found, it is true, that
some fillers give an ash lacking in compactness and liable to split
asunder, but the Rca to flake seems to be controlled almost entirely

«The terms light and heavy as eel in ae connection refer to the body or thick-
ness of the leaf, which largely controls the rapidity of the burn.

FIRE-HOLDING CAPACITY AND EVENNESS OF BURN. Ja

by the character of the wrapper and binder. Furthermore, this lack
of cohesion in the ash of a wrapper may be largely overcome by the
use of a good binder. As to the color of the ash, a binder having a
good burn will impart to the ash of the wrapper a lighter tint anda
more uniform color. The general results of these experiments may
be summarized as follows:

(1) In order to secure a good burn, due consideration should be given
to the proper balancing of the components of the cigar; thatis, a heavy
filler should be wrapped with a comparatively heavy wrapper, while a
light-bodied filler requires a light-bodied wrapper.

(2) Of the three components of the cigar, the filler exerts the strongest
influence on the evenness of the burn.

(3) The influence of the wrapper and binder is shown most strongly
on the character of the ash, and the binder very materially influences
the ash of the wrapper in this respect.

TESTING THE CAPACITY FOR HOLDING FIRE AND THE
EVENNESS OF THE BURN.

The factors of holding fire and of burning evenly are of prime
importance in judging the burn of tobacco, and any sample found
markedly deficient in these points may be rejected without applying
any further test. As has been previously stated, the cld method of
measuring the fire-holding capacity is likely to lead to erroneous con-
clusions, while there has heretofore been no direct method of deter-
mining the evenness of the burn of wrapper leaf. In the process
which we have used for testing wrapper tobacco with regard to these
elements of the burn, the area of the leaf consumed, rather than the
time elapsing before the glow is extinguished, is measured.

The form of the apparatus used in this method will be understood
by reference to the accompanying illustration (fig. 2). The essential
feature is the form on which the leaf is wrapped, consisting of a col-
lapsible wooden tube, one end of which fits into a glass tube of the
same diameter. This latter is in turn connected with a second glass
tube through which is drawn a current of air. The best material for
making the wooden form is well-seasoned cherry with a straight grain,
but ash has also been found to answer the purpose very well. From the
wood selected is made a cylinder 5 inches in length, ¢ inch in diameter
at one end, and tapering slightly to the other end (see fig. 3). In the
larger end of the cylinder a 2-inch hole is bored to a depth of 3% inches,
and the shell thus formed is separated into six equal segments by saw-
ing to a depth of 34 inches. The smaller end is cut down for a dis-
tance of 13 inches, so as to fit snugly into the glasstube. The shoulder
thus formed should correspond in depth to the thickness of the wall of
the glass tube. Near the larger end of the form a groove (c) is cut,
into which is fitted a rubber band. The plug (d@) has a diameter such

12 TESTING THE BURNING QUALITY OF CIGAR TOBACCO.

that when inserted in the end of the form the latter is expanded to its
original size. The receiver (a) for the form is made by drawing out

Fic, 2.—Apparatus for testing the burning quality of wrapper tobacco: a, entrance of air current; b,
wrapper to be tested; ¢, c, glass tube to which the form bearing the wrapper is attached by means
of the cork, m; d, d, large glass tube fitted with corks, n, n, through which passes ¢, ec; /, flask
containing water; 0, small glass tube dipping beneath the surface of the water in J; p, short glass
tube leading from /; h, pump by means of which the current of air is secured; e, e and g, g, rabber
tubing connecting parts of apparatus; f, water tap; 7, outflow of water.

one end of ashort piece of thick-walled glass tubing. Allof the above

dimensions are based on tubing having an internal diameter of 14 mm.

(9 16 inch) and an external diameter of 18 mm. (1116 inch). The

small end of the re-

ceiver is fitted with

a Sone cork ()). by

means of which it is

connected with the
other portion of the
apparatus.

From the leaf: to
be tested, which
should be quite

damp, the Wrapper Fic. 3.—Construction of form on which leaf is wrapped for use in

Ch : ae ne apparatus shown in figure 2: a, glass tube for receiving the form; },

its cut inte a form cork by which receiver (a) is connected with remainder of apparatus

quite similar to that shown in figure 2; ¢, rubber band for collapsing the form; d, plug
: for expanding the form; e, forr yhich leaf is w d.

used for cigars, and oO panding the G rm on which leaf is wrappe

the same rules are observed as regards the cutting of right-handed and
left-handed wrappers, ete. Beginning at the outer end the wrapper is

FIRE-HOLDING CAPACITY AND EVENNESS OF BURN. 13

rolled quite tightly, first on the form and then on the glass. At the
beginning of the process of rolling, the extreme outer corner of the
base of the wrapper is attached to the overlapping portion with a bit
of cigar paste, and at the end of the operation the tip of the wrapper
is attached to the receiver by the same means. A number of samples
to be tested are thus wrapped on the forms and set aside until they
have dried out properly. The plug in the end of the form is then
withdrawn and the rubber band causes the walls of the latter to col-
lapse, so that it can be easily withdrawn from the receiver. This
leaves the sample of wrapper securely attached to the glass tube,
and in exactly the same form it would have on a cigar. The tube
“arrying the sample to be tested is connected with the remainder of
the apparatus, shown in figure 2, the construction of which will be
understood without further explanation. The current of air is fur-
nished by means of an ordinary filter pump, and its rate can be con-
trolled with suflicient accuracy by measuring the flow of water through
the pump. The end of the wrapper is ignited with a flat gas flame,
and the evenness of the burnand the portion consumed before it ceases
to glow are carefully noted. Our method of recording the results is
to grade each sample on a scale of ten, both with reference to the
evenness of the burn and the fire-holding capacity. Of course, stand-
ards in these tests are purely arbitrary, as the results are only intended
to be comparative. Under the conditions laid down for the experi-
ment, wrappers having markedly good burning qualities will burn up
completely and evenly with only one lighting, and these are given a
grade of 10.

_ For the purpose of comparing the results obtained by this method
with those given by the cigar test with reference to the evenness of the
burn, a number of leaves were selected from different types of wrap-
per tobacco. One half of each leaf was used for wrapping a cigar and
the second half was wrapped on the form for testing, as has Just been
described. There was a decided lack of agreement in the results
obtained by the two methods when only one type of filler was used in
making the cigars. It was found that frequently a wrapper that
eraded only 5 or 6 ona scale of 10 in what may be called the ** form
test”? would burn quite evenly on the cigar, whereas another wrapper
grading as high as 9 in this test would show an uneven burn on the
cigar. A good illustration of this pomt is found in a wrapper which
was scored 10, 9, 10, respectively, in three experiments with the form
test and gave a fire-holding capacity of 65 seconds by the old method
of Nessler. On one type of filler this wrapper gave a very uneven
burn, but when smoked ona lighter filler the burn was perfectly satis-
factory. These results, then, seem to emphasize the fact that, although
the final judgment as to the burning qualities of a wrapper which has
shown up well in the preliminary tests must be based on the smoking

14 TESTING THE BURNING QUALITY OF CIGAR TOBACCO.

of the cigar, great care must be exercised to avoid the sources of error
in this test which have been previously discussed. The test should not
only be repeated with a single type of filler to avoid the effects of any
possible unevenness or other imperfections in the manufacture of the
cigars, but at least two different types of filler should be used, one of
these being heavy and the other light in the sense in which these terms
are used here.

TESTING THE BURN OF CIGAR-FILLER TOBACCO.

Testing the burn of a filler is a much simpler problem than is the
‘ase with a wrapper. The principal elements of the burn are the
evenness and the capacity for holding fire, and the character of the ash
is unimportant, except that it should be compact. The evenness of
the burn and the fire-holding capacity are best determined by using
the cigar test. In the case of filler tobacco the capacity for holding
fire thus refers simply to the length of time the cigar will continue to
burn after being lighted without being puffed by the smoker. The
effects of the binder and wrapper on the burn may be avoided by
making the entire cigar from the filler leaf to be tested. Another
decided advantage in making the whole cigar from the same tobacco
is that the aroma, which is so important in the filler, can also be tested
at the same time. In determining the fire-holding capacity it is only
necessary to light the cigar and test it at gradually increasing intervals
of time to find whether it has ceased to burn. It is, however, desir-
able to test the fire-holding capacity and the evenness of the burn on
separate cigars if sufficient material is at hand for this purpose.

O

Bul. 100, Pt. 1V, Bureau of Plant Industry, U. S. Dept. of Agriculture. PLATE |.

D | C

5S a a

VARIATION IN BURN OF WraApPERS DUE TO DIFFERENT FILLERS: A, FILLER GROWN
IN TEXAS FROM CUBAN SEED; B, FILLER GROWN IN OHIO FROM DOMESTIC SEED;
C, FILLER GROWN IN SOUTH CAROLINA FROM CUBAN SEED; D, AN IMPORTED
CUBAN FILLER.

The same sample of Sumatra.wrapper was used throughout and the binder was taken from the
same leaf as the wrapper in each case.

Bul. 100, Pt. \V, Bureau of Plant Industry, U. S. Dept. of Agriculture. PLATE II

VARIATION IN BURN OF FILLERS DUE TO DIFFERENT WRAPPERS AND BINDERS: A, B, C,
THREE DIFFERENT TYPES OF CONNECTICUT-GROWN SUMATRA WRAPPER ON SAME
SAMPLE OF CUBAN FILLER, BINDER BEING SAME AS WRAPPER IN EACH CASE; D, SAME
WRAPPER AND FILLER AS C, BUT CONNECTICUT BROADLEAF USED AS BINDER.

Art ap rac iiant oly

LIBR ARY

viii 9