TOBACCO.
596. Fermented and Unfermented Tobacco.—Samples of tobacco may reach the analyst either in the fermented or unfermented state. As a basis for comparison, it is advisable in all cases to determine the constituents of the sample before fermentation sets in. The analysis, after fermentation is complete, will then show the changes of a chemical nature which it has undergone during the process of curing and sweating. Only tobacco which has undergone fermentation is found to be in a suitable condition for consumption. In addition to the natural constituents of tobacco, it may contain, in the manufactured state, flavoring ingredients such as licorice and sugar, coloring matters and in some instances, it is said, opium or other stimulating drugs. It is believed, however, that opium is not often found in manufactured tobacco, and it has never been found in this laboratory in cigarettes, although all the standard brands have been examined for it.[614]
In researches made at the Connecticut Station it is shown that fermentation produces but little change in the relative quantities of nitric acid, ammonia, fiber and starch in the leaves, while those of nicotin, albuminoids and amids are diminished. This is not in harmony with the generally accepted theory that starch is inverted and fermented during the process.[615]
The nature of the ferments which are active in producing the changes which tobacco undergoes in curing, is not definitely understood. Some of the organic constituents of the tobacco undergo a considerable change during the process. Any sugar which is found in the freshly cured leaves disappears wholly or in part. As products of fermentation may also be found succinic, fumaric, formic, acetic, propionic and butyric acids.
597. Acid and Basic Constituents of Tobacco.—In unfermented and fermented tobacco are found certain organic acids, among the most important of which are citric, malic, oxalic, pectic and tannic. Of the inorganic acids the chief which are found are nitric, sulfuric and hydrochloric. Among the bases ammonia and nicotin are the most important. Ammonia is found in the unfermented tobacco in only small quantities, but in the fermented product it may sometimes reach as high as half a per cent. The presence of these two nitrogenous bases in tobacco renders the estimation of the proteid matter contained therein somewhat tedious and difficult.
598. Composition of Tobacco Ash.—The mineral constituents of tobacco are highly important from a commercial point of view. The burning properties of tobacco depend largely upon the nature of its mineral constituents. A sample containing a large quantity of chlorids burns much less freely than one in which the sulfates and nitrates predominate. For this reason, the use of potash fertilizers containing large amounts of chlorin is injudicious in tobacco culture, the carbonates and sulfates of potash being preferred. The leaves of the tobacco plant contain a much larger percentage of mineral constituents than the stems, their respective contents of pure ash, that is ash free from carbon dioxid, carbon and sand, being about seventeen and seven. The pure ash of the leaves has the following mean composition: Potash 29.1 per cent, soda 3.2 per cent, lime 36.0 per cent, magnesia 7.4 per cent, iron oxid 2.0 per cent, phosphoric acid 4.7 per cent, sulfuric acid 6.0 per cent, silica 5.8 per cent, and chlorin 6.7 per cent.[616]
599. Composition of Tobacco.—The mean composition of some of the more important varieties of water-free tobacco is shown in the following table:[617]
| Havana, per cent. | Sumatra, per cent. | Kentucky, per cent. | Java, per cent. | |
|---|---|---|---|---|
| Nicotin | 3.98 | 2.38 | 4.59 | 3.30 |
| Malic acid | 12.11 | 11.11 | 11.57 | 6.04 |
| Citric acid | 2.05 | 2.53 | 3.40 | 3.30 |
| Oxalic acid | 1.53 | 2.97 | 2.03 | 3.38 |
| Acetic acid | 0.42 | 0.29 | 0.43 | 0.22 |
| Tannic acid | 1.13 | 0.98 | 1.48 | 0.51 |
| Nitric acid | 1.32 | 0.60 | 1.88 | 0.23 |
| Pectic acid | 11.36 | 11.88 | 8.22 | 10.13 |
| Cellulose | 15.76 | 10.59 | 12.48 | 11.82 |
| Ammonia | 0.49 | 0.06 | 0.19 | 0.23 |
| Soluble nitrogenous matter | 7.74 | 8.84 | 13.90 | 10.39 |
| Insoluble”” | 9.75 | 7.97 | 8.10 | 9.53 |
| Residue and chlorophyll | 5.15 | 8.63 | 1.99 | 6.45 |
| Oil | 1.03 | 1.26 | 2.28 | 0.81 |
| Ash | 17.50 | 17.03 | 14.36 | 18.46 |
| Undetermined | 8.68 | 12.88 | 13.10 | 15.20 |
Among the undetermined matters are included those of a gummy or resinous composition not extracted by ether, the exact nature of which is not well understood, and the starches, sugars, pentosans and galactan.
Tobacco grown in more northern latitudes has less nicotin than the samples given in the foregoing table.
The following table shows the composition of tobacco grown in Connecticut:[618]
- (A) = Unfermented,
- (B) = Fermented,
| Upper leaves. | Short seconds. | First wrappers. | ||||
|---|---|---|---|---|---|---|
| (A) % | (B) % | (A) % | (B) % | (A) % | (B) % | |
| Water | 23.50 | 23.40 | 27.40 | 21.10 | 27.50 | 24.90 |
| Pure ash | 14.89 | 15.27 | 22.85 | 25.25 | 15.84 | 16.22 |
| Nicotin | 2.50 | 1.79 | 0.77 | 0.50 | 1.26 | 1.44 |
| Nitric acid | 1.89 | 1.97 | 2.39 | 2.82 | 2.59 | 2.35 |
| Ammonia | 0.67 | 0.71 | 0.16 | 0.16 | 0.33 | 0.47 |
| Proteids | 12.19 | 13.31 | 6.69 | 6.81 | 11.31 | 11.62 |
| Fiber | 7.90 | 8.78 | 7.89 | 8.95 | 9.92 | 10.42 |
| Starch | 3.20 | 3.36 | 2.62 | 3.01 | 2.89 | 3.08 |
| Oil and fat | 3.87 | 3.42 | 2.95 | 3.04 | 2.84 | 2.92 |
| Undeterm’d | 29.39 | 27.99 | 26.28 | 28.36 | 25.52 | 26.88 |
600. Estimation of Water.—In the estimation of water in vegetable substances, as has already been noted, it is usual to dry them in the air or partial vacuum, or in an inert gas, at a temperature of 100° until a constant weight is reached. By this process, not only the water, but all substances volatile at the temperature and in the conditions mentioned are expelled. The quantity of these volatile substances in vegetable matter, as a rule, is insignificant and hence the total loss may be estimated as water. In the case of tobacco a far different condition is presented, inasmuch as the nicotin, which sometimes amounts to five per cent of the weight of the sample, is also volatile under the conditions mentioned. It is advisable, therefore, to dry the sample of tobacco at a temperature not above fifty degrees and in a vacuum as complete as possible. Tobacco is also extremely rich in its content of crystallized mineral salts, containing often water of crystallization, and there is danger of this crystal water being lost when the sample is dried at 100°. The desiccation is conveniently made in the apparatus described on [page 22]. If a high vacuum be employed, viz., about twenty-five inches of mercury, it is better not to allow the temperature to go above 40° or 45°. A rather rapid current of dry air should be allowed to pass through the apparatus for the more speedy removal of the moisture and a dish containing sulfuric acid may also be placed inside of the drying apparatus. It is possible by proceeding in this way to secure constant weight in the sample after a few hours.
601. Estimation of Nitric Acid.—The nitric acid in a sample of tobacco is most easily estimated by the ferrous chlorid process.[619]
The sample is best prepared by making an alcoholic extract which is accomplished by exhausting about twenty-five grams of the fine tobacco powder with 200 cubic centimeters for forty per cent alcohol made slightly alkaline by soda lye. The mixture is boiled in a flask with a reflux condenser for about an hour. After cooling, the volume is completed to a definite quantity, and, after filtering, an aliquot part is used for the analytical process. It is evident that the nitric acid cannot be estimated in this case after previous reduction to ammonia by zinc or iron on account of the presence of ammonia in the sample itself. If, however, the amount of ammonia be determined in a separate portion of the sample, the nitric acid may be reduced in the usual way, by zinc or iron, the total quantity of ammonia determined by distillation, the quantity originally present in the sample deducted and the residual ammonia calculated to nitric acid.
602. Sulfuric and Hydrochloric Acids.—These two acids are determined in the ash of the sample by the usual methods. The sulfuric acid thus found represents the original sulfuric acid in combination with the bases in the mineral parts of the plant, together with that produced by the oxidation of the organic sulfur during combustion. In order to avoid all loss of sulfur during the combustion, the precautions already given should be observed. The separation of the sulfur pre-existing as sulfates from that converted into sulfates during the combustion is accomplished as previously directed.[620] For ordinary purposes, this separation is not necessary.
To avoid loss of chlorin from volatilization during incineration the temperature should be kept at the lowest possible point until the mass is charred, the soluble salts extracted from the charred mass and the incineration completed as usual.
603. Oxalic, Citric and Malic Acids.—The separation and estimation of organic acids from vegetable tissues is a matter of great difficulty, especially when they exist as is usually the case, in very minute proportions. During incineration, the salts of the inorganic acids are converted into carbonates and the subsequent examination of the ash gives no indication of the character of the original acids. In the case of tobacco, the organic acids of chief importance, from an analytical point of view, are oxalic, citric and malic. These acids may be extracted and separated by the following process:[621]
Ten grams of the dry tobacco powder are rubbed up in a mortar with twelve cubic centimeters of dilute sulfuric acid (one to five) and then absorbed with coarse pumice stone powder in sufficient quantity to cause all the liquid to disappear. The mass is placed in an extraction apparatus of proper size and thoroughly extracted with ether until a drop of the extract leaves no acid residue on evaporation. Usually about ten hours are required. The organic acids are thus separated from the mineral acids. The ether is removed from the extract and the residue dissolved in hot water, cooled, filtered, if necessary several times, until the solution is separated from the fat and resin which have been extracted by the ether. The filtrate is neutralized with ammonia, slightly acidified with acetic and the oxalic acid contained therein thrown out by means of a dilute solution of calcium acetate, which must not be added in excess. The calcium oxalate is separated by filtration, and determined as lime oxid. To the filtrate is added drop by drop, with constant stirring, a dilute solution of lead acetate, prepared by mixing one part of a saturated solution of lead acetate with four parts of water. When the precipitate formed has settled, the clear supernatant liquid is tested by adding a drop of acetic acid and a few drops of the dilute lead acetate. In case a precipitate be formed, the addition of the lead acetate is continued until a precipitate is secured which will immediately dissolve in acetic acid. At this moment the citric acid is almost completely precipitated. In order to avoid the accumulation of the acetic acid by reason of the repetition of the process as above described, the mixture is neutralized each time with dilute ammonia. The precipitated neutral lead citrate obtained by the above process, is separated by filtration and, in order to avoid its decomposition when washed with pure water, it is washed with a very dilute acetic acid solution of lead acetate. The washing and filtration are accomplished as quickly as possible, and the final washing is made with alcohol of thirty-six per cent strength. In the filtrate the residual lead citrate, together with a little lead malate, are precipitated by the alcohol used as the wash and this precipitate is also separated by filtration. The filtrate containing the greater part of the malic acid is evaporated to remove the alcohol and treated with lead acetate in excess. Afterwards it is mixed with five times its volume of thirty-six per cent alcohol containing a half per cent of acetic acid. In these conditions the lead malate is completely precipitated as neutral salt, and after standing a few hours, is separated by filtration. The three precipitates, obtained as above, are dried at 100° and weighed. If the precipitates have been collected on filter paper they should be removed as completely as possible, the papers incinerated in the usual way and any reduced lead converted into nitrate and oxid by treatment with nitric acid and subsequent ignition. From the quantities of lead oxid obtained, the weights of the citric and malic acids are computed. The precipitate which is obtained by the action of alcohol, above noted, is also dried and ignited and the lead oxid found divided equally between the citric and malic acids, the respective quantities of which found, are included in computing their total weights. The weight of the citric acid is calculated from the formula (C₆H₅O₇)₂Pb₃ + H₂O, and that of the malic acid from the formula C₄H₄O₅Pb + H₂O.
604. Acetic Acid.—For the determination of the volatile acids of the fatty series existing in tobacco, the following process, also due to Schlösing, may be followed:[622]
The apparatus employed is shown in [Fig. 12]1. Ten grams of the pulverized tobacco, moistened with water and mixed with a little powdered tartaric acid, are placed in the tube A. The two ends of the tube, A, are stoppered with asbestos or glass wool. Steam, generated in the flask, D, is passed into B. After fifteen minutes, or as soon as it is certain that the contents of A have reached a temperature of 100°, the dish, F, containing mercury, is placed in the position shown in the [figure]. The steam, by this arrangement, is forced into the lower end of A, passes into the condenser E, and the condensed water collected in C. The operation should be so conducted as to avoid any condensation of water in B. It is advisable during the progress of the distillation, which should continue for at least twenty minutes, to neutralize from time to time the acetic acid collected in C by a set solution of dilute alkali, or, an excess of the alkaline solution may be placed in C and the part not neutralized by the acetic acid determined at the end of the distillation by titration.
Fig. 121.—Apparatus for Acetic Acid.
605. Pectic Acid.—Under this term are included not only the pectic acid but all the other bodies of a pectose nature contained in tobacco. These bodies are of considerable interest, although they do not belong to the most important constituents. In fresh tobacco leaves are found three pectin bodies. One pectin is soluble in water, another is an insoluble pectose and the third is the pectose body forming salts with the alkalies, i. e., true pectic acid. In fermented tobacco pectic acid is found chiefly in combination with lime in the ribs of the leaves, serving to give them the necessary stiffness. For the estimation of the pectin bodies (mucilage) the powdered tobacco is thoroughly extracted with cold water. An aliquot part of the aqueous extract is mixed with two volumes of strong alcohol and allowed to stand in a well closed vessel in a cool place for twenty-four hours. The precipitate is collected on a filter, washed with sixty-six per cent alcohol, dried and weighed. The dried residue is incinerated and the amount of ash determined. In general, vegetable mucilages contain about five per cent of ash. If more than this be found, it is due to the solution of the salts of the organic acids contained in the sample. A dried vegetable mucilage, obtained as above, dissolves in water to a mucilaginous liquid which does not reduce alkaline copper solution until it has been hydrolyzed by boiling with a dilute mineral acid.[623]
606. Tannic Acid.—This acid is separated and estimated by the processes given in paragraphs [589-595].
607. Starch and Sugar.—The unfermented leaves of tobacco contain considerable quantities of carbohydrates in addition to woody fiber, pentosans, galactan and cellulose. Among these, starch is the most important. Sugar exists in small quantities in the fresh leaf, usually not over one per cent. During fermentation, according to some authorities, the starch is partially converted into sugar and the latter substance disappears under the action of the alcoholic ferments. It has been found at the Connecticut Station, however, that the starch content of the leaf does not decrease during fermentation. The starch and sugar may be determined in the fresh leaves by the methods already given.
In the manufacture of certain grades of tobacco it is customary to add a quantity of sugar. The analyst may thus be called upon to determine in some cases whether the sugar found in a sample is natural or added. The occurrence of natural sugars in tobacco has been investigated at the instance of the British Treasury.[624]
The natural sugars which may be found in sun dried tobaccos usually disappear entirely during the process of fermentation. It was found by the Somerset House chemists that the content of sugar in commercial tobaccos varies from none at all to over fifteen per cent. A remarkable example of this variation is reported in two samples from this country, one of which, grown in Kentucky, contained no sugar, and the other grown in Virginia, 15.2 per cent.
It was noticed that the saccharin matters in the tobaccos examined were neutral to polarized light. They are determined by their copper reducing power. The tobacco sugars are therefore to be classed with the reducing bodies, not optically active, found in the juices of sorghum and sugar canes.
608. Ammonia.—As has already been intimated, ammonia exists only in minute quantities in fresh tobacco leaves, but in considerable quantities after fermentation. In the estimation of ammonia, twenty grams of the tobacco powder are digested with 250 cubic centimeters of water, acidulated with sulfuric and after an hour enough water added to make the total quantity 400 cubic centimeters. After filtration, an aliquot part of the filtrate, about 200 cubic centimeters, is treated with magnesium oxid in excess and the ammonia and nicotin removed by distillation in a current of steam. The distillate is collected in dilute sulfuric acid of known strength. The total amount of the two bases is determined by titration and the quantity of base representing the nicotin, which has been determined in a separate sample, subtracted in order to obtain the weight of the ammonia.[625]
The ammonia in tobacco is determined by Nessler in the following manner:[626]
The powdered tobacco is mixed with water and magnesium oxid and after standing for several hours it is distilled in a current of steam, the distillate received in dilute sulfuric acid and the process continued until a drop of the distillate gives no reaction for ammonia with the nessler reagent. The excess of sulfuric acid in the distillate is neutralized with pure sodium carbonate and the nicotin precipitated by a neutral solution of mercuric iodid and potassium iodid. The precipitate is separated by filtration, the filtrate treated with sodium sulfid, and the ammonia again obtained by distillation with an alkali, collected in dilute solution of set sulfuric acid and determined by titration. The difference of the two determinations represents the ammonia.
609. Nicotin.—In this laboratory McElroy has made a study of some of the best approved methods for determining nicotin, and finds the most simple and reliable to be that proposed by Kissling.[627] The finely powdered tobacco should be dried at a temperature not exceeding 60°, or it may be partially dried at that temperature before grinding and the final drying completed afterwards. Twenty grams of the powdered sample are intimately mixed by means of a pestle with ten cubic centimeters of dilute alcoholic solution of soda lye, made by dissolving six grams of sodium hydroxid in forty cubic centimeters of water and completing the volume to 100 cubic centimeters with ninety-five per cent alcohol. The mass is transferred to an extraction paper cylinder, placed in an extraction apparatus and extracted for three hours with ether. The ether is nearly all removed by careful distillation, the residue mixed with fifty cubic centimeters of a very dilute soda lye solution (4 to 100) and subjected to distillation in a current of steam. The flask containing the nicotin extract should be connected with the condensing apparatus by a safety bulb as is usual in the distillation of substances containing fixed alkali. The distillation should be conducted rapidly and in such a manner that when 200 cubic centimeters of the distillate have been collected, not more than fifteen cubic centimeters of the liquid remain in the distillation flask. In the distillate, the nicotin is determined by titration with a set solution of dilute sulfuric acid, using rosolic acid or phenacetolin as indicator. It is advisable to titrate each fifty cubic centimeters of the distillate as it is received and the distillation is continued until the last fifty cubic centimeters give no appreciable quantity of the alkaloid. In the calculations one molecule of sulfuric acid is equivalent to two molecules of nicotin according to the equation
- H₂SO₄ = (C₁₀H₁₄N₂)₂.
- 98324
Polarization Method.—Popovici has based a method of detecting the quantity of nicotin in tobacco on its property of rotating the plane of polarized light.[628] The gyrodynat of pure nicotin is expressed by the formula [a]D = -161°.6. When ten parts of nicotin are mixed with ninety parts of water, this value becomes -74°.1. By reason of this great depression in gyrodynatic value Popovici determined the relation which exists between the dilute solutions of nicotin and the number of minutes of angular rotation produced on polarization in a 200 millimeter tube. In a solution in which two grams of nicotin are contained in fifty cubic centimeters, each minute of angular rotation is found to correspond to 6.5 milligrams of nicotin. For one gram in solution in the same volume one minute of angular rotation corresponds to 5.9 milligrams and for a half gram in solution to 5.7 milligrams.
The nicotin is prepared for polarization by extracting with ether, as indicated in the previous paragraph, and the ethereal solution from twenty grams of tobacco is shaken with a concentrated solution of sodium phosphotungstate in nitric acid by means of which nicotin and ammonia are precipitated and rapidly settle. The supernatant liquid is carefully poured off and the residue made up to a volume of fifty cubic centimeters with distilled water and the nicotin freed from any of its compounds by the addition of eight grams of finely powdered barium hydroxid. In order to promote the decomposition of the nicotin compounds the mixture should be shaken at intervals for several hours. The at first blue precipitate changes into blue green and finally into yellow. It is separated by filtration and the somewhat yellow colored filtrate placed in an observation tube, polarized, the polarization calculated to minutes of angular rotation and the number of minutes thus found multiplied by the nearest factor given above.
The analyst will find a description of other methods of estimating nicotin in tobacco in the periodical literature of analytical chemistry.[629]
610. Estimation of Amid Nitrogen.—For the estimation of amid nitrogen ten grams of the powdered tobacco are digested with 100 cubic centimeters of forty per cent alcohol, the extract separated by filtration, acidified with sulfuric and the albumin, peptone, nicotin and ammonia precipitated with as little phosphotungstic acid as possible. The precipitate is separated by filtration and seventy-five cubic centimeters of the filtrate evaporated in a thin glass or tin foil capsule after the addition of a little barium chlorid and the nitrogen determined in the residue. The nitrogen thus obtained is that which was present in an amid state. The nitrogen present as amids, ammonia and nicotin subtracted from the total nitrogen leaves that present as protein.
611. Fractional Extraction of Tobacco.—To determine the character of the soluble constituents of tobacco it is advisable to subject it to a fractional extraction with different reagents. The reagents usually employed in the order mentioned are petroleum ether, ether, absolute alcohol, water, dilute soda lye and dilute hydrochloric acid. The extract obtained by petroleum ether contains vegetable wax, chlorophyll and its alteration products, fat, ethereal oils, and resin bodies. The extract with ether may be divided into water soluble and alcohol soluble bodies. Among the first are small quantities of glucosids and nicotin while in the alcoholic solution resin predominates.
The alcoholic extract is also divided into water soluble and alcohol soluble parts. The first contains the nicotin, which is insoluble in ether, in combination with acids, together with tannic acid and allied bodies and also the sugar. The part insoluble in water consists chiefly of resin.
The aqueous solution contains the vegetable mucilages (pectin) soluble carbohydrates, soluble proteids and organic acids.
The dilute soda lye solution contains chiefly proteids.
The dilute hydrochloric acid solution contains the starch and the oxalic acid originally combined with lime. The extractions with dilute soda lye and dilute hydrochloric acid should be made at a boiling temperature. The residual matter consists of a mixture of carbohydrate bodies to which the term crude fiber is usually applied.
612. Burning Qualities.—When tobacco is to be used for the manufacture of cigars, or cigarettes, or for smoking in pipes, its ability to keep burning is a matter of great importance. The tobacco, when once ignited, should burn for some time and form, a fluffy ash, free of fused mineral particles. A tobacco with good burning properties is one containing nitrates in considerable quantity, not too much sugar and starch, a porous cellular structure and comparatively free of chlorin. In determining comparative burning properties the tests may be applied to the single leaf or the tobacco may be first rolled into a cigar form and burned in an artificial smoker.
Fig. 122.
Apparatus for Smoking.
In applying the test to the leaf it is important that the ignition be made with a fuse without flame, which maintains a uniform burning power. Any good slow burning fuse may be used and it is applied to the leaf in such a way that a hole may be burned in it, leaving its edges uniformly ignited. The number of seconds elapsing before the last spark is extinguished is noted. At the Connecticut Experiment Station a lighter, proposed by Nessler, is employed. It is prepared by digesting eighty grams of gum arabic in 120 cubic centimeters, and forty grams of gum tragacanth in a quarter of a liter of water for two days, mixing the mucilaginous masses and adding ten grams of potassium nitrate and about 350 grams of pulverized charcoal. The mixture is rolled, on a plate sprinkled with charcoal, into sticks a few inches in length and of the diameter of a cigar and dried at a gentle heat. These fuses burn slowly and without smoke and are well suited for lighting tobacco leaves. Several tests, at least six, should be made with each leaf. Leaves having a uniform burning power should be used as comparators and the number of seconds they burn be designated by 100. It is important that all the samples to be tested be exposed for a day or two to the same atmosphere in order that they may have, as nearly as possible, the same content of moisture. The burning tests, when possible, should be made both before and after fermentation. As a rule fermentation improves the burning quality of second rate leaves, but has little effect on leaves of the first quality.
613. Artificial Smoker.—For the purpose of comparing the burning properties of cigars, or of leaves rolled into cigar form, the artificial smoking apparatus devised by Penfield and modified in this laboratory is employed.[630] The construction of the apparatus is shown in the accompanying [figure.]
The lighted cigar is set in the tube at the left, so that air entering the test-tube must pass through the cigar. The test-tube contains enough water to seal the end of the tube carrying the cigar, and is connected with the aspirator on the right by the T tube, as shown. An arm of the T dips just beneath the surface of the liquid in the cup in the center. Water flows in a slow stream into the aspirator through the tube at the extreme right, forcing the air out through the arm of the T until the siphon begins to act. While the water is voided through the long arm of the siphon, air enters through the cigar, the liquid rising in the T. The action of the apparatus is automatic and intermittent. When the cigar is about one-third burned, it is removed without disturbing the ash cone, and the latter examined and compared with other samples as a standard. The sealing liquid of the long arm of the T may be mercury or water. In case mercury be used, care must be taken not to immerse the open end of the T more than one millimeter therein.