Tobacco Documents Online

4 Oriain of the cicarctte smoke. In contrast to a widesproad opinion, cigarette smoke is no a produat pS. comhustton. In the presence of oxygen and at high temporatures, com- bustion leads to degradation products which are rich in oxygen. In the cigarette, the tobacco itself does not undergo such a reaction; only the carbon in the leaf does this. Therefore, the actual combustion products in smoke are only carbon dioxide and carbon monoxide. The many hundred smoke constituents of an organic mixture originate during heating, practi- ca]1y in the absence of oxygen, and therefore may be divided into two main groups: 1. Distillation products: substances which evaporate from the tobacco to:.rard the rear of the glowing zone, and are distilled away without under- going a chomical change. To this group belong all of the substances which are found in the to5acco and in the smoke -- such as paraffins, waxes, terpenes, and many volatile acids and bGses, as well as aromatic substances, among the various classes of substances. The most important distillation product is nicotine; it is especially interesting because on account of its limited volatility (boiling point 247°C) it is alt-says distilled only in the great,er part, while the rest is pyrolyzed (see : ibure 1). 2. Pyrolysis products: substances which have originated by random combination or rccombination of molecular fragments which were formed by the destruction of difficultly volatile and non-volatile substances (above all polysaccharidcs and polyreptid.es). Corresponding to the great role plzyed by chance in these changes, the pyrolysis products embrace practically the entire irLmense assortment of substances in the lower iaolocular weight organic chemistry. 'lhat reraains aftor distillation and pyrolysis is practically pure I

LTV,".ATU!'.r (1) Wenuech A.: Z. Untereuch. Lebonemittel. 7!, 178-186 (1937). [R] ii'snuiah A.: D" Tabakrauch. Arthur (}eist Verlag, Bromen 1939. (Vergriffen, freundliobet. weieo von 1)r. :-k lCuhn, Wion, (k herhuwn.) [3] Pyriki C.: ISnr: inet. T.bakforooh. Draulen 2, 127-141 (1966). (4) Borbdy P.: Z. Priiventivmodizin, 8onderhoftr Die Toxikologio dee Tab.ks l! (198Z). (6] Keith C. H. und J. C. Derriek: Tob. 8oi. V, 64-91 (1941). (8) lVartrnan W.B., l:'.C.Cogbift und E.6.Narlow: Ansl. Chem: dl, 1706-1709 (1959). (7] NoJjmann D. und E.L.Wytulcr: J. N.t. C.noec Inst. i0, 47 (1963) (deeteoh in Beitc. suc T.b.lfor.chung 2. 61-88 (1903). f81'Bnears A.: Tob. rici. VII. 78-80 i19831. • [9] Ykkcry H. B. und 0.1V. PucAcr: J. Biol. Chem. i(, 433-541 (1929). •. [10] Laurens A.H. und T.O.NarrcJt: Ansl. Chem. ?0, 1800 (1968): . ' (11j Grob K.: Beitc. z. Teb.kforech. 97-100 (1981). (12) Orob K.: J. of t3.e Chromatog. (im Druck). (1'] Lipp O.:• Vortng Croupe Frrnls du Cers.w, P.rie, !0. Juni 1963. 14

. the method described by Grobii. The reproducibility of the measurements is naturally best in the fiberglass filter and in the absorption liquid.. On tho other hand, the determination in the butt is less exact, on the one hand because small deviations in the butt length are unavoidable, and on the other hand be- cause thc amount of nicotine present in the unloaded butt was determined on a parallel sample, and therefore is subject to the small but always present differences from cigarette to ci.garettc, as an additional source of error. In the calculation of the nicotine loss, final.ly, it must be taken into account that the error is the summation of all of the other doter:nin- a.tion5. Since the amount of nicotine absorbed in the "saliva" is strongly deppax3.ent upon the liqui.d surface and the duration of the contact, the flow rat^ of the smoke andd the vibration of the-glass tube above all have to be reproduced as exactly as possible. RP.si)lts anr) ca i.scnssion As tr,a.y be seen from Figure 1 and Table 5, there is a considerable difference bet~recn the individual types of tobacco, 1•rith respect to the amount of nicotine carri.edd over from the smoke into the "saliva.11 This difference is shot-m not only in the absolute measured quantity of nicotine, but also in the percentage based on thee nicotine contained in the smol:e comin" out of the cigarette, or upon the nicotine contained originally in the tobacco. Only about 1/10th of the total amount of nicotine in the smoke leaving the cir:irettes is taken up by the "saliva" in the case of smo',;e of lo~•r p?i (1iGht :'irri.n); on the ot.hcr hand, in the case of an all:a- line reaction of the siiol.e (Para6ni3y, Jap1r,c:c.e Burley) about 1/5th is taken up. The P'lrylandd tobrieco, traditionally very much liked in Srritzerland, takes up an intermedi.lte position: Althourh ot.h9r factors could also ~_

-15 - Type of tobacco Ratio of pyridino and pyridino dorivatives to nicotine Virgin Rhodceicn ................... .... , 1,1 :1 Virgin USA ........... ............... 0,26:1 Maryland USA ....... . ....... . . : . . . . 4.. 0,48: 1 Paraguay ................ .:.......... 0,08 : 1 Burley Japan ........................ 0,10:1 According to this, the nicotine cannot only be absorbed more easily from s:nokA which reacts alkaline, but the absorbod rniaturo, in addition, is in genAral loss contaminated with pyridine and pyridine derivatives. Associated with this circumstance is also the fact that the alkaline smoke contains a larger fraction of the tobacco nicotine, since on the one hand less nicoti.ne is held back in the tpbacco butt, and on thp other hand the loss from hnatin- arx•3 the side stream is rather smaller. Our determination that on the one hand a higher fraction of the to- :A=QUU nicoLine is lost at lov:er p?i values, and on the other hand the fraction absorbed in the tobacco butt is likewise o eater than at hic;ier pa, may be explaincd in the folloi•ring way: In tobacco +.-rith an acid reaction, the nicotine is more stronolJ held t.)ack b°cause of its lower vapor pressure; it therefor.e gocs over into the c smoke only at higher temperature. On +.he one hand this causes an increased enrichnent in the tobacco butt and on the other hand the higher temperature necessary for the transfer into the smoke also ~ives assistance to an in- creased decomposition by heat. In this manner, a hi ;her proportion of the pyridinc fraction would T,e favored in the smoke. The situation is con- sidera.bly simpler w:ien there in an artificial constant filter material in place of the tobacco butt (Lipp13). .. We thank iterr A. Cossy for carryin~ out the nicotine± d.c~ters~inltion:.

smoke the problom is generally of less si;,mificance, in our case wo were afraid of greater disturbances. As a consequence of the much higher vola- tility of pyridine, the proportion of this in,the mixture absorbed by the "saliva" would have to increase significantly. For this reason, we have investigated Frith the gas chromatograph the composition of the photo- metrically determined fraction (the so-called °total alkaloids"). Capi1- lary columns were used uith a "coating liquld" which we synthesized, whieh is discussed separately (Grob12 ). Figure 6 shows a corresponding gas chromatogram. to ... 8. ~ ' a . •7 • . . A . _f Ilp' .1 ISO' /85' M.IA. Fig. 6. Gas cflromatograni or the hises Absorbedd in Vlc ''s;iliva" frorh the smoke from Japanese ?w.•ley. Coluin: 50 m10.5 m7 stationlry p:iase• • polycthyloneimine; carrier C?.s: nitrogen; detector: flzTno ionization. .;.d.ontificd co:r,nonents: ].. pyri- .iine, 2. 2-:.,ettiiylpyridine, 3. 2,6-Jimethyl.pyri:iine, !F. 2-'°thyl- pyrid=ze , 5. 3-methylpyridine, 5. 3-ethylpYr i{7.ine , 7. 2,4-d i- Methylpyridine, P,. 2,3-dimothylpyridino, 9. 3-cyanopyridinP, 10. nicotine. P'one of the tobaccos investig-ated produced, in the fractions absorbed by the "sali.va", amotLhts of acconP^.nying alkaloids which were definitely nieasurc-%ble. Tha raVo of pyridine to its simple derivatives (methyl, rlimethyl, and ethyl pyri3ine and others) showed no graat variation. On the other hand, thc: ratio of the total pyridine fractIon to the amount of nico- tine was q»itn diffcrent, as may be seen from the ."ollowing surtu,iary.

Z. PFAVENTrr1-M. 9. 14-25 (1964) e . NICOTINE ABSOP.F'TION FROM CIGAItETTE SMOKE (Delivered in abridged form.at the "Seventeenth Tobacco Chemists' Research Conferenco- I.ontreal, September, 1963) by A. J. Artho and K. Grob Addrisses: Dr. A. J. Artho, F. J. Burrus & Company, Boncourt Dr. K. Grob, Organic Chemical Institute of the University, Zurich . Summary Tho quantity of nicotine ("total alkaloids") absorbed from cigarctto smoko by an arti.. ficial "saliva" was found to vary,consid- orably botwoon different tobacco types. Thie indicates that neither tho amount of nicotine originally presont in the tobacco, nor the amount present in tho condonsod "t.otal particulate matter" of the smoke, can be taken as a reliablo measure of the physiological strength of difforent ciga- rottos. rree nicotine is eaeuy absorbea by a buffored liquid such as the saliva. With highor pH valuo of the smoke, the vapor •, pressure of nicotine (at 37 ° C) is also higher, thus more free nicotine is availablo in tho vapor phase, and more is absorbed by , the "saliva". The surface area contacting the smoke (depth of inhalation) and tho duration of the contact must also be taken in consid- oration. Furthormoro, since a considerable , portion of the nicotino can be rotainod in , the butt as the smoke passes through it, and revaporiaod as the burning zone advan- ooa, the importance of the ~butt length i. ' readily understood.. . Thp posi ttnn of ni ccti.nA nrnonn the :%^!AkA cc,nstit.u~nt:In spite of increased rec.earch efforts, nicotine, out of the hur-drods of known substances in cigarette smoke, remains at present the only one which could be studied as an individu.ll and well-dofined substancfi rrith smoko producns in definite physiological effects. For other effects which

P human there are unfortunately still no Fnoific causative agents knorm. Indeed, exact relationships between sourcos and effect have been discussed again and again; however, the hypotheses always remain poorly supported. And also, for these hypotheses certain i:rours of substances must always be kept in view, never a single substance such as nicotine. To be sure, tobacco contains several alkaloids. However, in terms of quantity, other than nicotine they play no noteworthy role in the smoke of practically all varioties of tobacco. hF rQlationshi.p between threA nicotine nuant.itIc_s_: in the tobacco. in the smoke, and in the body of the smoker. It is much more important to have the most exact information possible concorning the nature and the mothod by which nicotine is transferred from the tobacco into the human body. For df:cades, the nicotine content of a tobacco has been considered as a reliable basis for the evaluation of its ..t,. . ~..~....z...1 . rr.,..~,. ~.rsas a~,. .,....i`.,1 ,r ....r3 r..~~........~~....d s:.,3L- .,1 .~..e t..~.~--'o_.....~ .~~ ._- -... .. ~~. -~ _~ r;.,...~ . .~....~r..,,_. we began to rely upon the nicotine content of the smoke -- an essential ad- vance, since the two values are connected in no simple manner. However, 1,2 Wenusch has shown as early as 1937 that even on tho basis of the smoke nicotine content, the alkaloid effect of a tobacco can be grossly misunder- stood! Later, Pyriki has also confirmed this thesis in several papers (for example, reference 3). Figure 1 shows that in this case we are not at all dealing with sub- tletiPs. Cigarnttos made from various crude tobaccos were smoked in a small tube, the inner wall of which was rinsed by p buffered liquid, which thus served as a model of the oral cavity (for details, see the discussion method). Amon~ the numbered examples used, the first two tobaccos produced an acid smoke, the last two produced basic smoke, while the X.aryland variety ~

-10- L A further difficulty results from the fact that the sucking in and expulsion of the smoke which occurs in naturstl smoking is very difficult to counterfeit mocha.nically. In the custorr,aty mechanical smoking equip- ment, a quantity of smol:e which corresponds to the dead volume of the absorption apparatus always remains in contact with the absorbing medium for the duration of the puff interval. Hot•rever, since in the present experiment the duration of the contact existing bet•reen tho smoke and the "saliva" is directly of decisive significance, we have done away with the intermitting of the smoke puff, and have smoked instead in a"eontinuous puff'' in A. flota of 80 ml/minute. The size of the glass tube was so chosen that at thP given flo!•7 rate the smoke •.ias in contact with the liquid svr- face for 2 secoM^. This type of snol-in- also hinders among other thin--s an cxcessive "ar-in-" of the smoke after the exit from the end of the cigarette (I:eith and Derrick5). As may be concluded from the light yellow color of the absorption liquid, the~.re i.s no notet•rorthy precipitation of the smoke particles in the tube, so that the quantity of alkaloid present in the liquid can be regarded. as pract~cally exclusively absorbed from the gas prase. r All ciglrettes rTgre snoked. to a butt lendtl: of 23 mm. Then the butt, the artificial "saliva" and the fiberglass filter were acidified with HCI, and the "total alkaloids" t•rere determined as nicotine:, spectrophotomFtri.- ca.lly, after stean disti111tion (Lai:rpne and Harre1110). The nicoti.ne ori.-in-i.17.y contained in the smoked tobacco ura.s determined on a cor:;ara.con serfes of ciC:rpttes. The Amount of nicctine; held back in the t•obacco butt •••as obt~ined from the difference beti•lePn the r+.nount determi.ned in t:Ie 0 zmokMl butt and thr- or;-in!?l• ar..eiint rrn.ent in the unsmoked butt. ThA p'i values of the smoke for the indivzdual types of tol--occo vere detormin-1 by V 0 N ~ ~ ~ t o* 0

! 1 VIBRATOR to We smoking machine ~ ~IBSs tube t•-ith 0.6 nS off buffer solution Fig. 3. Schematic repre:.c.ntation of the experimental arranre- mcnt for smokine. I s ri J. h. !'i.F!i•* o." the ~x~rim~r+.al rpparatts duping :moki.n,;. At talE Inrt, iS V,n c1.-1r^.•Ftc; ln th!o c!•ntor nf i!in r7.ct.nre a tlic VC•11ss tt>>)n t41A.r11 di.hn into t'in v1tcr ~t,d in r•ov,,d l-,y thn vibrator; on thn ri;,llt ;s tyc t'~r0,rid"o fi.]t.nr arIrl flnc :.ucticn connoction to tlin :melan„ mrlchir.o.

6 in the gas form disappears from the smoke because of absorption on a suit- able surface, the loss is immediately replaced by evaporation from the particles, And thus the disturbed distribution equilibriury is reestablished. This leads to a surprisingly rapid transport of nicotine from the particle phase to the interfaces of the surroundings, without the particles having to contact these interfaces directly. This transport mochanism was first established experimentally using phenols as examples (Hoffman and Wynder7; Spears8). The behavior of the nicotino.in smoke is decisively influenced by its characteristics as a base. In basic surroundinos, nicotine is present in the form of a neutral molecule, tinich could be found in the gas phase. In acid surroundings it is protonated, and has the forin of positive ions which could exist only in the liquid. or solid phase. As the neutral point (pII = 7), the nim;erical ratio of molecules to ions is about 1;4 (Vickery and Puchcr9). A decrease in the pH value signii'ies an increase in the num- ber of ions, and therefore a decrease in the volatility. As is apparent from Figur.e 1, the distribution of the nicotine may now be explained in the follo•-.in~~ l•:ay: Pyrolycis or di:,+.i114tinn: more nicotine is pyrolyzed from acid tobac- co, because it is less volatile. Therefore, it stays behind lon~er as the glowing zono advances, and is exposed to higher temperatures. Some nicotine is also always volatilized with the side stream, between the sr:o;:c puffs. - Abn,crp1 aon in the ar^). ewit_y: the sm~.^.ll surface of the oral c~vity can ho'..d bicl: onl.y a vlnishi.n;ly small portion of the smoke dro-~let.s. On the other hand, the saliva is an effecti.vo solvent,for the ni.cotinr: in the r; ~ S ~~,L ~.`^.. The saliv:, ;nay then withdraw n icotino from the r•~rticle }~hasA

. 13 - exist beWeen the smoking conditions which we chose, and the natural "intermittent" smoking: 1. As a consequence of the lack of smoking pauses, in the continuous puff there can be practically no loss of nicotine into the side stream smoke. 2. The rate of flow is constant in the case of the continuous puff, and corresponds to a flow rate of 1 1/3 ml/soc. In intermittent smoking, normally a volume of 35-40 ml is sucked in within 2 seconds, in the case of 1 or 2 puffs per minute. The x'loar rate is therefore 15 times higirer during a puff than in the case of the continuous puff; this has as a consequence the fact that the gloi•ring zone temperature is subject to greater variations, and higher ma.yimvm temperatures are obtained. Therefore, higher losses are to he expected from hoati.ng. NrnaovAr: A rrFatPr flow rate must also decrease the absorption in the tobacco butt. Actually i••e detPrminedd that in the case of intermittent smoking a con- sidcrably higher portion of the nicotine contained in the burned tobacco is lost, and that the aricunt absorbed. in the tobacco butt is significantly r smaller. The totp•1 nicotine leaving the cigarette in the main stream of smoke is like*•-rise somei•rhat smaller. On the other hand, we could also deter- mine differences in physical-cheTdcal behavior of the nicotine in the smoke which werc characteristic of the types of tobacco investigated, in the case of int,)r.m.lttent smoking. To be sure, Lreater difficulties are involved in the quantitative determination of these differences in the casa of inter- mittent smoking. . It is known to be difficult to determine the nicotine alon~•, in rni; - turos with pyridine and its derivatives. ~lhilc in tobacco and in the total

S carbon, which proves to bo present in the rogion where the oxygen is sucked in,'and combines completely with the oxygen, bringing about further dis- tillation and pyrolysis because of the heat oS combustion set•iree. It is rEn•rardins to emiu'asize this method of origin of the smoke, in order in this way to dispel the widespread error, according to which smoke should be made physiologically harmless by a "complete combustion." Structur.e of the cicarette smoke. In an invisible medium (the so-called gas phase) there are suspended as r. so-called particle phase about 3 billion droplets per cubic centimeter of space; these droplets have diameters of 0.1 to 1 11 (Keith and Derrick5). In the physical sense, these two portions are differentiated above all by their readiness for contact. Since free diffusion prevails in the gas phase, components in their forward motion in the smoke come quantitatively into direct contact with the bodies in their surroundings (pieces of tobacco, paper, filter fibers, saliva in the mouth, etc.) in a~.ztter of seconds. The droplets in the particle phase, on the other hand, behave very slug- gishly, and come into actual contact with the surroiindin~s only when these offer especial.ly large surfaces and a very fine subdivision of the smoke in the smallest space elemonts, such as is the case in an effective s.nol:c fil- ter or even in the lungs. In a chemical sense it has been shown'that by far the grAa.test nuaiber off smoke constituents are distributed in some ratio in both phases. Only a few substances with extremely low or extremely high boiling points are practically limited to one of the two phases. In Ennor- al, nore than °9p1' of the nicotine is in the particle phase6 , so thst for determination of the nicotine in smoke only the condensed particle phase is custorarily nmlyzr:d. f;owevcr, until recently, the .^.en:,itivity of the dis- tributien aquilibri>>m had been cons3.derably urn-lorestim~.ted: whon nicotino

3 took up an intermediate position. The behavior of the nicotine upon smoking proved to be quito dependent upon the pH value of the smoke. Most striking are the differences in the "oral cavity." There, the "saliva" took about four times as much nicotine from the basic smoke as from the acid smoke! T~bacco ype M Ilico- .tinR per.~ tobac-i co ( a e o, •a cot ,:e u:'i.,lt; srae ng ' ~ of nicotin ~onta ned in smoked ~, ,r,no .,`'r„e~, I le.rin stream t Vola- ~ rc-d r+ p?i of liz~ sorh~d e: - the ' r F}yro- ~. ine-~ . Tow ' n the by~ thn 1~'' Qd ~ -- butt, ivall~pA,,.ed s:no`ce • Virgin Rhod. 14,4 30 70 42 2,6 2s 5,0 Virgin USA 13,3 29 71 45 3.1 23 6.2 Maryland USA . 10.2 30 70 30 6,0 26 6,1 Psrsgusy 11.6 .6 94 34 13,2 .47 7,7' Burley Japsn . 27,4 '10 28 . 11,2 46 6.0 , . , . . . ~ ~.; ` ~. .~:....a...,~.: Fig. 1. Percenta6e distribution of tobacco nicotine among its ~... _ t _t .. T, . . ! ! 1 .' _ , 9 • YQl1~l4J ~lr,r1C:~lvil,r,.V PiNVt V:IC vu,vA.J.Ai6• •L1G 1,YSli1 111bV16dN 11GJ in the relationship between the pH value of the smoke and the nicotine absorption in the saliva of the mouth. The nicotine v:hich had passed the oral cavity, in the case of !'_cPr inhalation, was taken up to the extent of over 90o in the lung (Borbely 4 ). In this case, practically thc entire nicotine from the smoke Goes over into the body of the smoker, so that the smoke nicotine content is of a very certain and direct value as evidcnce. However, inhalation is a relatively young custom, and hazardous to health; here we are proceeding from the assumption that tt:e smoke is not inhaled.. In the caso of the two basic tobaccos this is also scarcely possible, since the basic smoke produces a painful irritating effect in the larynx. . For the understanding of the measured values, some information concern- ing the origin and r.tructure of the smoko is necessary.

- 12 - influence the vapor pressure and thus the "absorbability" of the nicotine in a buffered medium, we have, on the other hand, found the pronounced dependence upon the pFI of the smoke to be confirmed in the case of all of the investigations carried out thus far. Table 5. Amounts of nicotine ("total alkaloids") per cicarette. Average values from 3 determinations, in each case on one ciga- rette. In the last two columns, the three individuEl, measure- monts are also given. (P'cr percentage distribution, see Fig. l.) i I Years Types of t smo}:ec ,os -. zn ,o-tin^ soroc in the so,• e In :e . in t.~:e I sma!ce, tobacco tob~.cco nd siae tobacco saava conden- St:ri bU Bt . S t 3 ~g m m g mg ~ Virgin 0,32 3,48 Rhodesien 0,33 ' - 3.66 0,40 ' 3,23' 13,8 4,2 6,8 ' 0,35 3,45 Virgin USA 0,33 2,61 0.28 2,16 ' 0,38 2,60 10,6 3,1 4,8 0,33 2,42 Maryland USA " n.n ~ a! 0,42 1,91 0,38 1,40 6,7 2,0 2,8 0,l0 1,68 Paraguay ' 1,21 4.64 1,08 ; 3,98 1,32. 4,17 9,1 0,65 ' 3,1 1,20 .. 4,23 Burley Japan 2,6G ' 11,03 2,12 10,24 3,34 11,20 23,8 3,7 8,8 2,87 10,82 eC~o, '.-1Prnischh had esti:-r~).ted the fraction of the tobacco nicotine taken up from the smoke in the case of mouth smoking as 2-5~ from "acid" ci~arottes on the basis of an entirely different type of inea5urement. The values for Virgin found in our model experiments (Figure 1) agree amizinZly With this. In the interpretation of the exporimenta.l results, obviously the fact must be taken into consideration thot above a11 the followint- difforences

7 via the previously mentioned disturbance and reestablishment of the dis- tribution equilibrium, without actually coming into direct contact with it. I'7aturally this effect durin7 the short time oS residonce of two seconds assumes agroater importance, the more volatile the nicotine in the smoke particles is, i.e., the higher the pH value of the smoke ist .Absorntion in the ci;arettc butt: here the butt is understood to be the unburned tobacco residue, with no filter. A considorable portion of the smoke particles are held back on this tobacco, on account of its great sur- face. In add.ition to this, the tobacco fibers also influence thc smoke via its ~a.s rha.se. From this, we i-rould have to conclude that basic smoke also leaves behind more nicotine in the butt than does acid smoke. However, tho contrary is the case, on grounds which will be discussed later. ExnPri.^,~nt_1 ~c`hn~3s • fn,• niir i mrACt.i na t.i nrG wrrA n7•Eh11'ed on an auto- no tic lUac'1L''1~ ('•~olins '::~rk VIII). Len~th: 70 mm; dianeter: 8.3 m;7. I'orr each type of tobacco, the average cigarette weir,,,ht was determined after prior conditionin- at 60;' relative humidity and 20°C. 1'ie used. exclusively cigarcttcs i*nic'a varicd by less than 20 n- from the average weig-tit (techni- cal d.at.1, TKble 2). Ta.b7e..'.. Tec%nical data on the ciCarett.es investiCated. The duration of bun:,in•, is b•+sed on thF srro;ces' len,-th of 47 mnl at a flou of 80 m1 por rlinutc. TJlls iJ~ , w.~urat~-r;j watn° cc:n. !1vv tr- :0tt af Y,r,. nir. • t0'rt of ~ °l,r/ci.~^.rct Virgin Rhodoeian 1428 10,0 13,0 Virgin USA ` 1189 9,4 12.2 Tlaryland USA 948 6,4 ~ 11,2 P6raSuay 1173 8,9 11,3 13urloy Japan 1281 9,0 11,6 i

8 The Rr.porimental arranCemnnt for the sr,okfno m1y be tseen from Figures 3 and 4. Tho mouth piece of the ciCarctte is fastened to the ci.&,arette holder by an airtight connection 1•rith thin latex mambrane ("dontal dar,;" ). An a.ntor~^tic smol:in- ihachine of the "cigarette components CSM 10" type guarantees reproducible suction conditions. Ths fresh s:noke passes first thNough a Elass tube held at 37°C in airr.~ter bath, in which there is 0.6 nil of an absorption liquid whi.ch i:nitates the ::aliva of the mouth. This arti- ficial "saliva" conaists of 0.25 ?t phosphate buffer 4rith the addition 'af 100 p;+n of sa,wni n. The p•FI valu° was adjustod to 6.5, which corrosponds to the zvPra;;e va'!vc of ?^.uian saliva. ."\f.t-;r the c,at from the g11^s tube, the SmO~\E' ~nJi ^s through fi.nc filter ("CaTb:•id~e Filter"), An whi.c hch over 90.;,1f? of the particlcs cf (`.3 '1 c?i cr.;cter cont1ined in t'ie smol:e are held baclc6. This "smoke ~condea- sato" ("totn.) rsrticulat~ natter") which is precipitated on the fibcrgla•s : filtcr contains aa:on;; ether thin;;s practically all of the nicol,inE3 from t}:o s^ro:'c nz:sin,- throu3h it. In thc case of mouth smol:i.n~ (without inhalation) the sn:ol~e is : ucked into the nouth, t•here for one to t~-o seconds ;.n contact with the rr.oist mouth surface, and is then F:::pelled again. BetrTa.^,n successive smok: puffs there irs a rinsin~ of the r~outh sv.rfaces by r!;nct•ral of the saliva. Thc total 1^:cur.t of SUrf1co in t.ho oral cavity, as i•rell as its rinsin~ bet!rren sieccrssi.ve s•noke puffs, is probably susceptible to stron~;1•y indi- vidna1 vr.riltior.:>, •^•nd is di.ffi.cult to ricasure. In order to takc into r.ccowit those t-ro rlctor: in our c;;perimo:.t, in spi.t.c ot tizis, vc have ur.cc' a viL-ratur (50 vibrat,ion: p~,r second) to move the tLbe with the buffer solution. In this vrav the surface of thc artificial saliva i.s Lrer>tly in- (ir.-ivc forPl3tion) an,-' Zn addition :.s contin»zll;,~ rcnc~•rert.