Tobacco Documents Online

THE EFFECT OF FILTER COVERAGE OF CIGARETTE VENTILATION HOLES UPON SMOKE DELIVERY William L. Dunn, Ph.D.;1 Sidney Weinstein, Ph.D.;2 Willie Houck, B. Sc.;1 Frank Ryan, M.S.;1 Jos~ Nepomuceno, B.S.1 and Curt Weinstein, M.A.2 1Philip Morris Research and Development Center, Richmond, VA 2NeuroCommunicati'ons Research Labratory, Danbury, CT Correspondence and reprint requests to: William L. Dunn, Ph.D. Philip Morris Research and Development Center P. 0. Box 26563 Richmond, VA 23261

The placement of fingers on cigarettes while smoking was recorded~t?)113 '~~ smokers by three concealed video cameras.~#e~smoked their usual brands (venti- + lati on type cigarettes) under in situ conditions. Three judges independently &Xfi^`b determined from analysis of slow motion playback of the video tapes that ~ 41% of the puffs taken, fingers overrode the band of holes for at least part of the puff. Finger coverage occurred during 24.1% of the summed puffing time for the mean cigarette. The incidence of coverage was unrelated to FTC tar delivery It was further determined from controlle6labora- tory studies that when fingers deliberately overlaid the band of holes, 43% to 60% of the hol!es were covered (representative brand means). FTC tar delivery increments resulting from such coverage were also determined. From these find- ings were calculated the net effects of the observed finger coverage upon FTC tar , deliveries of the representative brands. The FTC tar delii'very increases for the . mean cigarette of each brand studied which coulid be attributed to the observed /w0 incidence of finger coverage of the holes were i'n ~)_Kinstancel~r6 greater than IF- 0.5 milligrams.

N which deliver less than <-~ `14 mg FTC tar embody some variant of the so-called "ventilated" filter. Minute holes penetrating the wrapper (tipping paper) of a conventional filter allow The Effect of Finger Coverage of Ventilation Holes a ir to be sucked into controlled amounts of ambient ~ ~ the cigarette to /erge with 4~ F .the mainstream smoke from the mout end of the cigarette. A smoke-control system W~~ L-I which combines the effect of the ventilation holes ww~w1 '~d1-' with the effect of the conven- n ~1 C- tional cellulose acetate filter affords the manufaturer an excellent means for liz reducing the typical 25 mg of FTC tar generated'by the burning coal to any de- sired delivery level within the 1-15 mg range. Kozlowski et al. O recently reported observations of smokers blocking VeN these ventilating holes with lips and fingers, Ufts:,-increasing the smoke delivery levels above the levels which would have been delivered had the holes not been 11or blocked. Seventeen of their 39 subjects (low-yield cigarette smoke5) were ob- served to block the holes "to some extent." Six of the 17 were blocking with their fingers.

he.4 continuation of the inquiry now requires that ~~ further observations be made of smoking behavior in order to determine the prev - ~ ~ Ao„uA 1'ence and the extent of blockage during actual puffing time and" this #0 P` blockage incidence into ~ alteration in the amount of tar~and )del i vered ae to the smoker. Such is the purpose of the investigations we now report. The report is based upon two iri;dependent sets of observations. Neuro- Communications Research Laboratories conducted the behavioral investigations in which naive smokers were unobtrusively videotaped~for the purpose of recording the occurrence of hole coverage by the fingers. This study is reported' as Section I - Behavioral Observations. - The secon6 set of observations were carried'out in the Product Development Directorate of the Philip Morris Research & Development Center. These consisted of determinations of the extent of hole blockage when fingers override the band . of holes, and systematic smoke measurements which established the relationshp between incremental hole blockage and the resultant change in smoke delivery for a group of representative cigarette brands. This study is reported as Section II - Ventilation Hole Blockage and Smoke Dellivery. R . .. •~ a :. '.ti. ~ . ,„f,t C.-

These two investigations are being reported together in order to integrate the results of each into a final statement - an extrapolation of observed hole blockage effects into calculated real world smoking effects (Section III - The Effect of Finger Coverage Upon Smoke Delivery).

L #V Section I - Behavioral Observations Subjects (Ss) A total of 137 subjects were selected for participation in this study. They ~ ti ~ were recruited from the Danbury, Connecticut metropolitan area and environs by .I newspaper advertiisements,~~ouncements posted in local stores and the state college campus,~nd from NCRL's general subject pool. The recruitment copy made no reference to cigarette smoking, asking only for paid participation in adver- tising research. were screened'on initial tellephone contact by means of questions about their age and their consumption of coffee, soda pop and ciga- rettes. Anyone smoking less than a pack a day, those outside the age range of 18-55, and those who reported their regular cigarette brand to be other thani those brands with unfilled quotas at the time of interview were rejected'. v 5s were assigned to one of three groups in accordance with the FTC tar delivery level of their reported regular cigarette brand. Twenty-four selected SS "'s were subsequentlly rejected for the following reasons: a during test (12), smoked brand not listed (9), used e,. uwvMI.akU ~ record inadvertently destroyed A 3). The us~able d~i dqot smoke '~ n cigarette holder i) or video 113 Sf ~ sa+b3eE-bs were about divided among the three groups and the two genders in each group (:see Table _ for specific numbers). were about equalTyy equallyl represented 1 „ ~?' l ~ ~~I ; ~

Procedure The participant arrived at NGRL to be greeted by a receptioniist and seated in a small waiting room where a sign was posted which read: "Please Do Not Smoke I I 5 0- Here." This was done to reduce the possiblity that ~ticaQaui would light a ~ all, s.w1" cigarette off-camera. The qualifications of t.`.c -#ee= were rechecked during i the $e&s-ion an&the subject-qualifying formiwas completed. Fhe-p4FtiQ-iperrt was led into a TV viewing room where the cameras were posi- 5~ Oe1D a`~,Bt~l A° ' --~ tioned as described'below: The pa~a~rt 'was'told to watchn&~ normally A SS dvne at home, and that smoking was allowed while watching TV. The s~ were also told that their responses to TV viewing would be monitored but no mention was made of video-recording. No ~d'eE-t asked questions regarding video- -KIC-'!r' recording prior to the completion of the testing session. If ^°~T asked le- e about the purpose of the study they were told that the investigators were ( -n . ~_ ~ l intereI-s_te"dn in TV viewing behaviors and responses to advertising. Itt,? `hr'~ A+4Y1 FwMN`3 .A X9&01I &ntP~t0 ad dA~ 2Dr\9 ~I ri CvnrPanbe& ,tUr~. CalwwPlLC~al~~¢~~~o:. ~p a~:a~c,4t ~~An~dhtrcaoy cotaining the butt and ashes of a previovusliy smokedicigarette was ~ . A g" placed next to the subject. If tJie~~ did not light a cigarette after several minutes, an experimenter~~nt,ered the room and asked ~ if 1gRat S L QwSwfn tu6.4 ~/ usual ly smoked while watching TV. If the 4ub3ee*--"M}+ "no," >:ko-e+rperimertter

left the room and waited 5 to 10 minutes more before re-entering the room and asking i;he-5th4es4 questions about adverti sing. If t*&--ozbjer&t said "yes" but cigarettes wtW=Aizw, made available a pack,80-1 c,tP"t-+eS- o., ~g E P- qk I own brand. Upon completion of one cigarette tj;f-, WA `tomfL 74.e 9 Q1va~ e waited a few minute s, then re-entered the room afiO questionad taa- A Orbgae#_: None of the subjects were told to smoke or otherwise influenced to ~. ~ S smoke beyond the above descr ption. If asked "Do ~ a.$Dyl-~CQ 1~ZtO1,4iutiA Ots~ cuXLi E. ~ A~- you want me to smoke?" il r responded by saying, "As you wish." After the session eachza-4ee-t- signed a release form andApaid t4Qor-agreed fee by the receptionist. The cigarette butts were collected, labeled, and stored. Equipment Three video cameras were positioned to obtain a front, side and top view of each sVbjett-. Two of the cameras were placed outside the testing room concealed in a speaker enclosure (front view) and lighting fixture (top view). The side 49 view camera was placed on its side on a table in the testing roomn with other pieces of photographic and laboratory equipment that appeared to be in teporary 4-- storage. The top and side views were fix-focused on the . ~~,.:. .. WA" Zoo328ss'7s f-~- #}~ , _; .. :. . __...,:,.., ~. ,.

0:_ stationary`armless chair. The front view of the sg6p~t was adjustable by # GdM&CACo„ ~ operator in the adjacent room. 2:ac~ &U;¢~yG4 o-v~ A'f.j,~ 4o ~..ca &Vn1 Vi &0 Me'r_L s~~,c~. , . ~ 0 ao 0 the video tapes. The tapes were played back at a speed that required 10.6 seconds of play-back time for the A passage of one second of real recorded time. Although the judges all viewed the tapes at the same time, they were visually isolated~rom each other by partition 4!~- : r panels, and procedural rules assure6independent judgements. The tapes could be stopped and segments rerun by an operator upon request. Judges were specifically instructed to record dubious instances of finger contact and hole coverage as actual finger contact and hole coverage events. Frequencies and durations of puffs, hole coverages with fingers~and finger . Ss contacts for all s=b4e-&ks were recorded by each of the judges. The durations eJkpA..a-lcff; ~ were recorded Ain play-back time, then converted to real time for reporting. The results of these analyses are reported in the form of two indices A , of finger hole coverage: 9W'Au^ I

. The Incidence Index - This index is based upon straightforward frequency ' counts of the observed instances of finger coverage occurring on a puff-by-puff _ ;i.-..,....V',.. basis summed over the smoking of the entire cigarette, and is reported as a percentage of the total puffs taken on the cigarette. It is an inflated measure ~since finger coverage during any fractional part of a whole puff was recorded as ;,coverage for the total duration of that puff. ~OV A/ first appearance of co~1 flare-up to the cessation of co~l flare-up, or remval of A the cigarette from the lips, whichever occurred first. Puff durations were summed across the series of puffs to yield the total puff duration for the ciga- rette. When finger coverage occurred on a given puff, coverage duration was recorded for that portion of the puff duration during which the fingers were in CwfilatYhk1d)tc-"~ - contact with the cigarette. These puff-by-puff times were summed across the V series of puffs to yield the Duration Index, expressed as Oercentage of the total puff duration. The Duration Index - Puff.-duration is defined as the time interval from the

~ ~ .. ha~.. t70 pnr^f Iu~JO 9'~F19VbJ,''iSpnF},;:~(} a9Of16f2ni G°~1 ~2do 9c . :.. . . .: : . :.. ; . - ,. , sraw'2'rioifb-rub`~ju9.; J2'i11 b9-t'rUJJU .f9V9(~Jff{MJ i2(,rf sr!3 a,o7i siie.ifp 6pf a_ 9fi3 lol no rJs -1u5r #fiuq f r 1o3 rsf!1 t`r f 5 iV-.03 zt I uq- fio 29 i^cQ2 yriJ r ,.+1, . ,2sw.noff,,-yvti 4psi4voa ~3fiuq eravrs, eT rio 54~r7uo:,r; y};5 t9v:;o 79tanr3 f~-6M: .~: 1s r. ` s.^rsw 2.-1Spnfl~fif` fiO~filN 7uEi tirn'h-tub liuq qr?:1 le ro~3-Yoq ff5ri.t 7ol ti9t` ro~e7 .,.'. .4 .. ~ ~ -/. •~1,-. ~ ~. , jT ~~+ ~~.r. ~ q ; j .4 ..~. i~~ }rr + ^ ~ :S - fJ f .._.., ,,,pJ 3f ... 9 b~ ~~r r ~•9 tct} t~G ~R~lU2 .' 9 r.3W 2~ Iri t +~~ u4-~;V- 1J(~. _Z. : . x . . t. . ~92Z~1'tqXJ ~k9brt1 fi0fi4f 7LC' q,If 519l~ od ?I~UC? 3c: 29J 7S2 . r,zc,i fE,-ruh 1}ur FS.ioj

Incidence Index- ~Hole coverage by the fingers at some time during the puff was judged to occur in 41% of the 1,105 puffs recorded for 113 subjects smok g their own brand of cigaretfe. . Women displayed a higher Incidence Index than men (44.7% vs 36.2%) but the e~ i Sl ~1 t { Q ~ d , t ux Y e s M l du difference ~ort~jr-aNr~ina~.ly significanGQ(t = 1.71, d'.f. = 111, p < .09). The , ~ .. . ~. appeaunrelated to the delivery level of the cigarette smoked ` e ~S ~ Y ~A~ ~ a d ~ Gmtiw7.u.~~~1L . V ~ '7~ ' Duration Index-- The Incidence Index isbased upon straightforward frequency A counts of puffs for which finger coverage occurred at any time during the puff. Since most smokers interrupt finger contact momen- tarily while puffing7 a more accurate representation of the degree to which fingers block the holes is the proportion of puffing!time in which the hole- covering fingers maintain contact withithe cigarette. This measure is the Dura- 6u_~5 tion Index, obtained for each subject by dividing the summed i^..~'~~ of hole '~ ~: (67 coverage ~ by the summed puff durations. His value, averaged across Cs- ~ all 113 subjects and expressed as a percentage 24.1% (S.E.M. = 2.0). I W OV~ ~ e77 i~~ll `

C, There is no difference between men and women ~ the Duration Inde " and difference relatable to the deliivery leveli of the cigarette being smoked. The s. .r. Duration Indices by group and'gender are shown in Table Di'scussion / (Insert Tab1L~ here) These observations suggest that the incidence of ventilation hole coverage f (as herein defined) may actually be greater than that repoted by Kozlowski et al. ~ (3). In fact the 15% frequency of hole coverage by the fingers reported by those investigators is to be compared with the Incidence Index of 41% which we have observed. Any number of factors could have contributed to~this di'fference, most likely prominent among them being the conditions under which the smoking oc- curred, and the greater opportunity to detect hole coverage on slow motion video playback from three angles.

The difference between the Incidence Index (41%) and the Duration Index (24.1%) reflects the frequency with which smokers release their finger hold upon ~. the cigarette during the puffing event. ~ appears to be a remarkably prevalent behavioral phenomenon. Among tKe 1,105 recorded puffs in this study, 802 (73%) were accompanied by release and regrasp finger motions. Some of these releases were quite brief, others spanned the entire duration of the puff. These observa- tions are of major significance in any rigorous evaluation of the exten6tl which ~ finger coverage of ventilation holes influences ultimate smoke delivery to the smoker. The hypothesis that smokers learn to effectively increase the amount of smoke they get from a ventilated'lcigarette by blocking the holes would have us expect to find a higher incidence of such blockage aong the ultra-low brand A l smokers. It is evident from Table L that such is not the case, with the absolute percentages trending in the opposite direction.

Section II - Measurements to Determine the Effect of Ventilating ,, Introduction Hole Blockage Upon FTC Tar Delivery Smokers do not block all of the ventilating holes when their fingers overlay the section of the cigarette cylinder containing those holes. The holes on the sides remain open. Therefore a closer approximation to real world events would based upon empi'rical determinations of FTC tar delivery increases that occur as a consequence of the partial hole blockaqe finqer coveraqe. f,.. . . . hole blockage from which can be generated the curve representing the tar , the band of holes and 2) The FTC tar delivery values for selected percentages of Two sets of values are needed to determine these approximations: 1) The observed extent of hole coverage for a sample of smokers when the fingers directly overlay delivery/hole blockage relationship. The Cigarettes We made these determinations for a variety of Philip Morris (PM)~bran6s selected to cover a range of dilution,levels and perforation methods (columns 1-

3, Table Z). These levels and methods are typical of those usediby other manu- A ''facturers and other brands. Experimental Procedures ;The panelist was then instructed to hold in random sequence cigarettes from each the usual cigarette-holding hand of each panelist were wetted with water colors. ,.:.,., ., To make the first set of measurements we convened a panel of 31 smokers from among PM R&D personnel in Richmond, Virginia. The middle and index fingers of f the brands in test as they normally hold a cigarette but with the constraint - P~ du~ ~ that th eirnwetted fingers be positionedAover the marked-off area of the ventila- tion holes. Th resulting water color print on each cigarette was then carefully blocked with glue and the dilution level for that treaed cigarete was measured e\ A using a standard PM Digital Dil'ution Instrument (DDI). The mean hole blockage percentage for each brand and the corresponding standard error appear in Table To generate the curve representing the tar delivery/hole blockage relationship for each brand in test, two or more levels of hole blockage were 1E i . , .f.Fl~. .. ,

selected for each brand and forty cigarettes at each blockage level smoked using n a PM 20-port smoking machine and standard FTC smoking methodology Hole blockage was achieved by applying glue over the prescribed number of holes prior o smoking. For 1~w Brand A only the two extreme ventilation levels were deemed necessary to test since the level of dilution for this brand is only 10% ( 10% of the total puff volume enters via the ventilation holes). Interpolated levels would alter the delivery level so slightly that the change would be ob- scured by the intrinsic error in measuring FTC tar delivery levels. At the *A note on methodology - Experience obtained in developing the dilution systems a for PM production cigarettes had'estblished that use of the DDI for measuring ~ A m dilution level provided the A st accurate means for determining tWopen area ~~ summed across all ventilation holes. Counting the number of blocked holes in this study would have been less accurate since the holes do vary somewhat in size, especially in the case of electric spark perforation. Fortuitously, use of the DDI is also the more rapid method. Changes in dilution level translate directly (though not necessarily rectilinearly) into hole blockage percentages.

other extreme, Brand E is designed to deliver smoke that is 80% diluted with - ,F ..~. ;•.. y+, ventil ation hole air. In this case the FTC tar delivery levels were measured at four levels of ventilation (95%, 60%, 30%, 0%)1. The remaining three brands fell (production dilution level) and the approximate midpoint. These 21 points are between these extremes. In these instances delivery levels were determined for three dilution levels; total hole blockage (0% dilution), no hole blockage plotted in Fig. 1'. Each set~'of points representing a given brand is fitted with the appropriate theoretical curve. The last column in Table-I is the FTC tar delivery value determined by entering the appropriate graph in Fig. !I at the mean finger coverage percentage level obtained for our group of 31 smokers and reading off the corresponding FTC tar delivery level. These values (last column, Table.L) represent in our judgment the best approximation to what happens to the FTC tar delivery of each brand when the smoker places his fingers directly over the band of ventilation holes. They represent the "worst case" delivery change because we have made the following conservative assumptions in deriving them: :.~~ . , ._ . _... .~.,.....,..~•..~.,.~,r.~.; .

1) f~i~ger coverage of a hole results in 100% blockage of that hole. f r ' !Ld- /h'04w; hdw-e ~n dLP Qccd2acc,q f teCp lOr7~ ,, Coverage constitutes perfect alignment of fingers over the band of :. ho l e s . V Finger coverage is continuous throughout the cumulated puffing time for th at c i g ar et te &6q--~cC~ mc 4YA 4w&-hA s T Ple ww Air; ~ ~ yez, kw Aea~~ ~~

Section III. Extrapolation to In Situ Smoking The FTC tar delivery increases appearing for each brand in the last column of Table ~ are increases which would occur were the hole-occluding fingers held 'in continuous contact with the cigarette throughout the entire time the smoker was puffing on the cigarette (mean puff time multiplied by total number of puff s). The NCRL observations reported in Section I make it evident that these values need be corrected. Under conditions comparable to the NCRL experimental conditions, the smoker is more likely to juxtapose fingers and'lholes for only a fraction of the puffing time and only a fraction of the occasions. Our purpose in this section is to integrate the findings presented in Sections I and II in order to arrive at reasonably rigorous and representative approximations to the changes in FTC tar delivery values which would likely ~6va=;~ from finger coverage of the ventilation holes during in situ smoking. (Insert Table:!~here) ._-;1 :,i.... r1: O O W N UZ m ~

Table ~ Percentage of Blocka e when Fin er s Cover Holes and the Effects upon ar e ivery by ran Number4- Rows of Perforation Hole s Laser Laser Laser jElectric Spark Perforation 2Based on data from 31 smokers % FTC Tar Blockage ---Deliveries (mg/cig)--- by Fingers2 Finger (S.E.M.) Unblocked Blocked 1 60 16.3 16.5 (3.8) 1 55 10.2 10.7 (2.0) 2 48 7.9 8.7 (1.8) 2 43 4.5 7.0 (1.6) 4 51 1.0 2.5 (1.5) t :

~.;; .. Column E in Table G represents the best estimate by brand o FTC tar delivery values (Column D) adjusted for the expected level of finger 1 coverage of the ventilation hg4es.' A comparison of Colis. D an&E will reveal a U5[C i.. ~N4~s~..~ most remarkable fact: None of the pubivi~'FTC tar delivery valuestare affected IK b6vl~"~S~..q drC EX+1VsId by the corrections for the brands evaliuatedL The FTiC values ape-pnli-}-i~Hed in whole numbers (.5 is rounded upward to the next whole number). Although throughout our investigations we deliberately erred in the conserv- ative direction so as to develop the "worst case" situation, there is a marke& disparity between our estimates of the net effect upon FTC tar delivery and the estimates reported by Koslowski et al. O~. Some discussion of this disparity is in order. Whereas Koslowski and his associates took into account all of the smoking behavior variables which could influence delivery, we confined our attention exclusively t'o hole coverage by the fingers. Interestingly, using the Koslowski ~ criteria for finger coverage, we observed 41% such incidence comle&to their 15%. The net influence of this effect was markedly redutd by two further modi- A fiers: 1) our documentation of the widely practiced finger release phenomenon, -/-4 .wcosfi,~ 4, t~ -VTQ 4-a-~-Au&wLeAS iLTLA7Cd (.t 2Qc~Q ,~s~~tp4 C~ lQB/)wertP 1o~ ~, . - ~ GUy nnuut4-r w>~,it o~ tj~~ 0 I I . `

the fingers being removed fromithe cigarette while iit is being grasped by the lips; 2) the partial! blockage of the holes by the fingers, even when the fingers are perfectly aligned with the band of holes. There are three other behavioral sources of potential delivery increase which Koslowski et al. considered which we did not: ,N S.Q.w-t ,~ ti .e'r, 0 of theAventilation system; 2)Acompensatory increases in puff volume and number of puff s; 3) coverage of the ventilation holes by the lips. Th first two sources will be at the disposal of the smoker w e ~ the ventlti llat 0 n s stem i s inco of the smoker to o rette, th',en it is t Cove age of the vent whatever limita ride the d 0 e exp porated int the cigar tte. If i is i tent ~livery red ction devic s de igned into the ciga- ~ ted a ions may be that he wi l disc er w ys and mea s to circ m- ~nn 0 sed b entilation h G)XF the can occur inadvertentl noted that hole cov and nwittiingly by the g P ly excl usi~phenomena. tv e zone 1 im en 0 pI~"~¢~,t~e~+~e'~`6t~ s ~~1 ~ fingers are ~;mutual- mm from the mo' th f the c'garette can be conceived'.as being the inimal covera A i e manufactu er. ips iis anoth mokerA. Howe by the lips4and hole co era 1) deliberate circumvention ~ r mattergsince this ~ er, it is by the ~~,, ~ pted by band~, 12 and ~'B \ zone; move-

1003285692

~....~.,......~.

:4_735V~ .r c me the rett~-usagel i affords , as is ev r, -that, t 0 1 f% ~ a rv~ : ' ~` , ery~here ~10fo oles outsidelthis zone in of anatomi alp~cclusio ~ yit~s~.~~ ven within his minimal dent f i net effec oportio un4ttainable the manufiac-turer a 1 occurring' in yr~e overaQje,zone so ~, . ~ e finding! s resente_d'in thi 3 i •. `. ., r . del i very `of .\s ~ 7, . ~1, . ... te a'major defi 4titude in e controlled re > "j J ; i a marketab e igarette. ither axial -ditection Areatly increases I I i i

J: 2~aasi~nii xfjsqtip\ otj3t~lib fsixt; 7«i,rig rri -6pf- )yJ )Gn 3o.t f !~,fy`Cr' i nor ur 3 ~: c, t_va, ~inos 2ir' J a:`cc z9fccf 9ri3 "I Jr, r .\tsl U6r;irll~~ 2rf + nt,v,: 1qpetu 9'.lf9~l 93atI!Q; rrJ !?J 25 <'n(, l:-+OaolQ dJw2 t` Jnn 2I f,4 tn l; ns7 a i rij finr riyU2 nOLSJ:`a`r~. ~flfry±nW-- / 3 / .°d,=l:rry5fcs . . ~ h I; r~, Ilrjf .,'.t~.f, 1.:-'CU f.iC'G.S9r1•r`,(,.•.n, : :) ir*•'-t_ r5 ?o Y'Srl ydeclo7rj 9rIJ' /6 h"j , .

REFERENCES '',`.1 2. k • Fa . .. . .i~u.'Sr 1 Federal Trade Commission (T979). Report of 'tar' and nicotine content of the smoke of 176 varieties of cigarettes. Federal Trade Commission, U.S.A. 2. .. . Al`.;;.•i *~ y .. . .. .. Kozlowski, L. T.; Frecker, R.C.; Khouw, V.; and Pope, M. A. (1980). .The misuse of 'less hazardous' cigarettes and its detection: Hole-blocking of ventilated filters. Amer. J. of Pub1ic Health, 70, 1202-1203. ~ Kozlowski, L.; Rickert, W. S.; Pope, M. A.; Robinson, J. C.; Frecker, R. C. (1982). Estimating the Yield of Tar, Nicotine and Carbon Monoxide from the .'lowest yield' ventilated cigarettes. British J. of Addiction, 77, 159-165. ," -- ,::I .;,. •