Fibre transfer experiments onto car seats

SCIENTIFIC & TECHNICAL

Fibre transfer experiments onto car seats

C ROUX*. J CHABLE and P MARGOT Institut de Police Scientifique et de Criminologie, University of lausanne, UNZL - Batiment de Chimie, 1015 Lausanne, Switzerland Science & Justice 1996; 36: 143-151 Received 6 February 1995; accepted 4 October 1995 Experiments on the transfer of textile fibres involving various garments, car seats, drivers and driving times were carried out in order to study their influence on the transfer of fibres to the seat of a car after it had been driven. Many fibres were transferred from the driver's garment to the car seat, even when the contact was brief and the garment and/or car seat was of a smooth texture. It was concluded that although all the parameters studied had a significant influence on the transfer, the garment was the determining factor. Therefore, in routine investigations, car seats should always be searched for fibres, but any assessment of fibre shedding from the garment will give only a preliminary indication of the numbers of fibres likely to be transferred.

Zur Ubertragung von Fasern auf Fahrersitze von Pkw's sowie ihr weiteres Verhalten bei der nachfolgenden Benutzung der Fahrzeuge wurden Versuche durchgefuhrt unter EinschluB verschiedener Bekleidungsstucke, Fahrzeugsitze, Fahrer und Fahrzeiten. Selbst bei nur kurzen Sitzkontakten und Kleidungsstucken bzw. Sitzbezugen rnit scheinbar glatten Oberflachen wurden viele Fasern ubertragen. Obwohl alle berucksichtigten Parameter EinfluB haben, kann doch geschlossen werden, daB letztendlich im jeweiligen Bekleidungsstuck der entscheidende Faktor fur das. ubertragungsergebnis liegt. Fahrzeugsitze sollten immer als potentielle Spurentrager angesehen werden. Die bloBe Abschatzung der Faserabgabekapazitat eines Bekleidungsstuckes fuhrt nur zu einem sehr ungenauen Hinweis auf die Anzahl der tatsachlich ubertragbaren Fasern.

Des expkriences de transfert de fibres textiles I? partir de plusieurs habits, sibges de voitures, chauffeurs et en fonction d'un temps de conduite, ont kt6 menCes afin d'itudier leur influence sur le transfert des fibres sur le sibge de la voiture, aprks que celle-ci ait 6t6 conduite. De nombreuses fibres ont CtC transfCrCes de I'habit du conducteur au sibge de la voiture, mCme lorsque ce contact Btait bref et que l'habit et/ou le sikge de voiture avaient une texture lisse. On a pu conclure que bien que tous les param2tres CtudiCs avaient une influence importante sur le transfert, l'habit Ctait le facteur determinant. C'est pourquoi les sibges de voitures devraient toujours Ctre soigneusement fouillts pour la presence de fibres, lors d'investigations de routine. Mais toute estimation de la chute de fibre du vCtement ne donnera qu'une indication preliminaire du nombre de fibres qui pourraient Ctre transferees.

Se llevaron a cab0 experimentos sobre la transferencia de fibras textiles de varios tipos de tejido, asiento de coche, conductores y tiempos de conducci6n, para estudiar la influencia de estos factores en la transferencia de fibras a1 asiento de un coche despuCs de haberlo conducido. Se transfirieron muchas fibras desde la ropa del conductor a1 asiento del coche aun cuando el contact0 fue breve y 10s tejidos de la ropa y/o asiento eran de textura suave. Se lleg6 a la conclusidn de que, aunque todos 10s parametros estudiados tenian una influencia significativa en la transferencia, el tejido era el factor determinante. Por lo tanto, en investigaciones de rutina, en 10s asientos de coche se debe siempre buscar fibras, per0 cualquier valoracidn del desprendimiento de fibras del tejido dara s610 una indicacidn preliminar del ndmero de fibras que seran presumiblemente transferidas.

Key Words: Forensic science; Criminalistics; Fibre evidence; Fibre transfer; Significance of fibre evidence.

* Corresponding author Department of Chemistry, University of Technology, Po Box 123, Broadway, Sydney, NSW 2007 Science & Justice 1996; 36(3):143-151

Australia

143

Fibre transfer experiments

Introduction Because of their widespread use and microscopic nature, textile fibres are some of the most readily transferred contact traces which obey Locard's exchange principle, that a criminal leaves traces of his passage at the scene of a crime and takes away traces of the scene. Although this was first formulated in 1920 [I], only since the mid-seventies has much research been carried out to verify the transfer theory and to try to acquire some knowledge of the phenomena which occur when two fabrics come into contact [2-61. These studies are important because they have led to a better understanding of transfer and persistence mechanisms involving fibres. The accumulated knowledge allows the expert to evaluate the number of fibres collected in the circumstances of a given case. Two questions occur at the time of the evaluation, regarding the transfer, persistence and recovery of fibres, if the suspected source was in contact with the receptor, and if the suspected source was not in contact with the' receptor. The answers to both questions, combined in a Bayesian framework, will improve the interpretation of the results of a forensic examination [7,8]. Most of the published data relate to garment-to-garment transfer, but many other situations will also generate fibre transfer, especially offences involving vehicles. For example, as quoted by Grieve [9], these may include sexual assaults or kidnapping, where a person has been transported against his or her will, robberies or terrorist attacks, where a vehicle may have been used for escape from the scene of the crime. Few studies have been reported and the data on the subject are limited.

In 1986, Robertson and Lim [lo] studied the transfer of fibres following a four-minute contact between four garments (donor) and four seats (recipient), as well as the subsequent persistence of the fibres transferred. They noticed that the number of fibres transferred was dependent on the donor and recipient used, and on the individual wearing the donor garment, more fibres being transferred from woollen garments than from those composed of acrylic fibres. They also found that the vinyl car seat used was a poor recipient of transferred fibres, and that more fibres were transferred to the vertical part of the seat than to the horizontal part, although this pattern was less marked in those cases where fewer fibres were transferred. Then, in 1989, Grieve et al. [6] included some transfer and persistence tests on car seats as part of their series of experiments related to the transfer of acrylic fibres, and confirmed some of the results obtained by Robertson and Lim. The present study is an extension of this earlier research work. It was carried out to study the influence of the driving time, the garment, the seat and the driver on the number of fibres to be found on the driving seat of a vehicle after use; to find indicators which would allow this number to be estimated; and to evaluate the validity of these indicators.

Experimental Material

Four series of experiments involving fibres transfer onto car seats were carried out. Eleven garments (donor) (Table 1) and 4 car seats (Table 2 and Figure 1) were selected, representing different fibre types, texture, construction and age. Ten drivers took part, as described in Table 3.

TABLE 1 Garnnents used as donors. No.

Garment

Composition

Construction

Link*

Texture

Age

8 x 15

smooth

new

14 x 29

smooth

2 years

1

Polo shirt 'J-J Benson'

cotton 100%

knit

2

Jeans 'Levis'

cotton 100%

woven

3

Sweat-shirt 'Ton-sur-Ton'

cotton 100%

knit

9 x 10

smooth

8 years

4

Pullover 'Colors'

cotton 100%

knit

1x 2

rough

3 months

5

Pullover 'Basic collection'

wool 100%

knit

6x9

rough

new

6

Jacket 'Benetton'

wool 100%

knit

3x 4

rough

2 years

7

Pullover 'La Pull'

wool 85%lpolyarnide lo%/ angora 5%

knit

6x8

rough

new

8

Jaquette 'TSE-Cashmere'

cashmere 100%

knit

8x7

rough

1 year

9

Pullover 'Hey'

acryl60%lpolyamide 18%/ mohair 15%lpolyester 7%

knit

3x 3

rough

2 years

10

Pullover 'Impulse'

acrylic 100%

knit

2x2

rough

2 years

11

Polar pullover 'Mountain Wave'

polyester 100%

knit

N.A.

rough

1 year

* [whaleslcm] x [courseslcm] for knitted garments, and [threads in warplcm] x [threads in weftlcm] for woven. 144

Science & Justice 1996; 36(3): 143-151

C ROUX, J CHABLE and P MARGOT

TABLE 2 Seats and vehicles used as recipients. No.

Vehicle

Seat composition Texture Age(years)

A Porsche 911 Carrera 2

leather

smooth

4

B Ope1 Corsa 1.2

polyester

smooth

9

C Toyota Camry GLI polyester

felt-like

8

D Renault Clio GT

hairy

1

wool

TABLE 3 People used as drivers. Driver

Sex

Weight (kg)

Seat A

Height (cm)

I

female

51

160

I1

female

55

163

In

female

58

167

IV

male

62

172

V

male

67

177

VI

female

71

165

VII

male

78

180

VIII

male

85

180

M

male

91

179

X

male

93

186

Methods Transfer experiments The back of each car seat was divided into 12 vertical strips, each 3.3 cm wide and about 250 cm2 in area (Figure 2). The seat was carefully cleaned with an adhesive brush before each contact and with a vacuum cleaner between each series of tests. Strict precautions were taken to avoid contamination between trials and to exclude the possibility of including fibres which had not come from primary transfers.

The driver wearing one of the test garments sat down, fastened the seatbelt and drove in town for 0,6, 12 or 25 minutes. For the 'zero-minute contact time', the driver sat in the seat, fastened the seatbelt, put histher hands on the steering wheel without starting the engine, then took off the seatbelt and finally got out of the car. Fibres were recovered immediately following the experiment using a 25 cm length of Sellotape Clear adhesive tape (Cellux AG, Rorschach, Switzerland), a new tape being used for each contact area. In order to preserve the microtraces collected, the tapes were stuck on to a transparency film for a plain paper copier (PP2410, 3M, Cergy-Pontoise, France). The tapes were then scanned under a Wild M7A low power stereomacroscope (Wild AG, Heerrbrugg, Switzerland) at a Science & Justice 1996; 36(3): 143-151

Seat C

" a

;-

Seat D FIGURE 1 Macro-photography of seats A, B, C and D (1 grad. = lmm) 145

Fibre transfer experiments

n

Estimation of the sheddability of the donors

The sheddability of the donor garments was estimated by placing an adhesive tape about 20 cm long on each donor at a random site. A weight of two kilograms of surface area 5.0 x 5.4 cm2 was then pulled with a string at a constant speed along the length of the adhesive tape. The experiment was repeated ten times at ten different places on each garment. The fibres present on a length surface of 7.5 x 2.5 cm were counted in situ under magnification.

FIGURE 2 Division of car seat back into twelve areas for collection of transferred fibres.

Estimation of differential shedding As donors 7 and 9 were of mixed composition, an attempt was then made to count the various fibre types individually. Three 4 cm length pieces of tapings were randomly selected for each driving time. They were then cut, mounted on a microscopic slide, and observed under brightfield microscopy at a magnification of 100x.

magnification of 12x using a scanning grid [ll]. Finally those fibres which were indistinguishable from the donor and whose length was greater than mm were counted in situ.

Transfer 'ffibres to the car seat The distribution of transferred fibres on the various marked areas of the car seats was examined after 0 and 25 minutes contact.

Four transfer experiments were carried out, as summarized in Table 4. Experiment 1, which varied the driving times, using driver VII, all donor garments, and the whole seat area, was designed to derive information about the influence of the driving time and the donor on the number of fibres found and their distribution on the seat. Experiments 2, 3 and 4 were carried out with no driving time. Experiment 2, again using driver VII and all donor garments, on seat areas 1, 6 and 11 in one car, was planned to show the influence of the donor on the number of fibres found, and the possible correlation of this number with the sheddability of the donor garment. Experiment 3 was carried out, using driver VII, three garments composed of different fibres and the selected seat areas in all cars, to demonstrate the influence of the recipient on the number of fibres found. Experiment 4, to indicate the influence of the driver on the number of fibres found, used all ten drivers, wearing one test garment, in the selected car seat areas in one car.

Results Experiments 1 and 2 - the effect of using different fibre donors and of varying the driving time The results of experiments 1 and 2, using different fibre donors and varying the driving time, are summarized in Tables 5 and 6 and Figure 3. In general, the number of fibres found on the back of the seat was high, even after a short contact; the simple action of sitting on a seat generated from just over a hundred to more than ten thousand fibres (Table 5). This number was influenced by the kind of garment and by the driving time; both factors were significant at 0.01 level with the variance analysis carried out, but the donor was the more important. The longer the driving time, the more fibres were transferred, but the increase was usually low in proportion to the number of fibres transferred immediately on the first contact. This increase as a function of the driving time seemed to follow a logarithmic progression; for most of the garments used, saturation was reached within 25 minutes (Figure 4).

TABLE 4 Summary of transfer experiments. Experiment

Driver

Donor Garment

1

VII

1 to 11

B

0,6, 12 and 25

2

whole seat

2

VII

1 to 11

B

0

10

areas 1 , 6 and 11 (see Figure 2)

3

VII

4, 6, and 9

A to D

0

10

areas 1 , 6 and 11 (see Figure 2)

10

areas 1 , 6 and 11 (see Figure 2)

146

Recipient Time of Car Seat contact [min]

Number of Su$ace considered trials per contact

Science & Justice 1996; 36(3): 143-151

C ROUX, J CHABLE and P MARGOT

TABLE 5 Experiment 1: transfer of fibres to car seat after various driving times (N=2). Donor garment

No offibres transferred after 6 mins 12 mins 25 mins

0 mins

TABLE 6 Experiment 2: transfer of fibres to areas 1, 6 and 11 of the car seat after a driving time of 0 minutes (N=10). looooo~

Donor No. of fibres transferred Gannent Mean Range GD

1

179

138-219

28

2

457

374-592

71

3

287

229-446

66

I

1

In

loo00

4

1001

+ - . - - - I

100

5

10

15

20

25

Driving time (min) FIGURE 3 Summary of the transfer of fibres from all garments (logarithmic scale).

Experiment 3 - the effect of using dzfferent recipients Experiment 3, using different recipients, is summarized in Table 7 and Figure 5. It appeared as if the nature of the recipient influenced the number of transferred fibres for each of the donors used (significant at 0.01 level with the variance analysis camed out). However, on further examination, it was seen that if the poor donor (pullover 9) and the poor recipient (seat A) were excluded, the number of transferred fibres was not influenced by the nature of the recipient. Science & Justice 1996; 36(3): 143-151

04

I

0

5

10

15

20

Driving time (min) c ~ ~ of fibres (garment 5). FIGURE 4 T v D ~transfer

25

Fibre transfer experiments

TABLE 7 Transfer of fibres from donor garments 4,6 and 9 to areas 1 , 6 and 11 of car seats after a driving time of 10 minutes. Donor

Car seat B

A

C

D

Donor 4 mean range kSD Donor 6 mean range ?SD Donor 9 mean range kSD

A

B

Seat

C

D

FIGURE 5 'kansfer of fibres from (a) donor garment 4; (b) donor garment 6; and (c) donor garment 9; lower and upper limits correspond to a standard deviation (N=10)

Experiment 4 - the effect of varying the driver The results obtained from varying the driver in experiment 4 are summarized in Table 8 and Figure 6. In this experiment the number of transferred fibres depended on the driver wearing the garment (significant at 0.01 level with the ~arianceanalysis carried out). The interpersonal differences were, however, mall. On 45 combinations of transfers taken two by two, 24 were indistinguishable(not significant at 0.01 level with the least significant difference statistical test carried out). In general, the number of transferred fibres remained constant, independent of the wearer of the garment. Furthermore, the number of fibres was not correlated to the weight of the driver. Additional information A closer examination of the results showed that the differences between two garments (donors) of the same composition could be more important than the differences appearing between two garments of different composition. For example, for two donors in 100% cotton, the number of fibres observed could vary from 800 to 5000 while it was possible to find two garments of a different composition that transferred approximately the same amount of fibres. This was the case with the 100% cotton shirt (garment 2) and pullover 7 (70% wool, angora 20% and polyamide 10%) which showed comparable results. 148

In the sheddability test, shown in Table 9, average values (N = 10) of shed fibres correlated positively with an average number (N = 10) of transferred fibres at t = 0 minutes (R = 0.771 and R2 = 0.595). The correlation was excellent (R = 0.988, R2= 0.977), if the smooth cotton garments, shirt 2 and sweat-shirt 3, which deviated from the regression line, were excluded. As would be expected, transfer results with donors of mixed composition showed that differential shedding occurred. With donor 7 (Figure 7), the percentage of angora transferred fibres was higher than the value given on the garment label, and the percentages of polyarnide and wool fibres

TABLE 8 Transfer of fibres from donor garment 7, worn by all drivers, to areas 1,6 and 11 of car seat B after a driving time of 10 minutes. Driver

I

No offibres transferred Mean Range

GD

530

415-687

73

5 14

333-671

106

I1 I11 IV

v VI VII VIII

IX X

Science & Justice 1996; 36(3): 143-151

C ROUX, J CHABLE and P MARGOT

Drivers FIGURE 6 3 a n s f e r of fibres from garment 7 worn by all drivers; lower and upper limits corres~ondto a standard deviation.

TABLE 9 Results of sheddability test. Garment

:I

No of fibres transferred Range Mean

1

160

2

2900

3

2059

4

1785

5

85

68-107

13

6

122

81-188

30

7

330

2 19-477

78

8

405

273-544

110

9

59

43-88

15

10

3724

11

82

2550-5540

1008

Garment

.

Driving time (min) FIGURE 7 Differential shedding of garment 7. Angora Polyamide W Wool B 100

90 80

&

70 60

B so 40

49-149

29

were lower. With donor 9 (Figure 8), the percentage of wlvester fibres was similar to those shown on the label. * . The percentages of mohair and polyamide fibres were higher, whereas the percentage of acrylic fibres was lower. It cannot be assumed that the shedding capacity of a garment will be related to the appearance of its texture. For example, shirt 2, which was smooth, shed many more fibres than pullover 9, which was rough, on the same recipient surface. Knowledge of the fibre composition or of the textile construction, used independently, will not allow an accurate estimation of the shedding capacity of a donor. Concerning the localization of the fibres on the seat, it was noticed that these were transferred to the whole of the seat (on the 12 collecting areas) whatever the garment. However, the distribution noticeably differed when the examination was made after 0 and 25 minutes of driving. In general, after 0 minutes there were more fibres on the door side than in the middle or on the gear side (Figure 9a). but Science & Justice 1996; 36(3): 143-151

I.

30 20 10

0

h

12

Driving time (min)

. 25

Garment

FIGURE 8 Differential shedding of garment 9. Polvester Mohair IPolvamide Acrvlic B

after 25 minutes driving, even though fibres remained relatively poorly distributed in the central area compared to the peripheral areas, distribution had reached an equilibrium between the left and right hand sides (Figure 9b). Discussion This study gives some answers to the questions which arise in so vast a subject as fibre transfer. Errors in counting were minimized because the contrast between fibres and seats were quite good; a control sample of the target fibres was used as an aid to recognition during scanning; and the number of background fibres coming from garments other than those considered were certainly very low (if not negligible) as the seats were rigorously cleaned before

Fibre transfer experiments

Car seat area FIGURE 9 Distribution of transferred fibres on the car seat after driving times of (a) 0 minuites and (b) 25 minutes.

experiments. The limitation in fibre length was chosen only for practical reasons (a more rapid scanning time allowed for more experiments and more samples). It does not mean that fibre fragments shorter than 1 mrn would not occur or would not be of value in a case situation. The methodology adopted leads also to several questions. It could be asked whether the fact of using perfectly clean seats did not facilitate the transfer conditions. One cannot reject the hypothesis that on a dirty seat, the transfer areas already occupied by unknown fibres make each new transfer more difficult. In other words, the absolute values encountered here could be overestimated compared to normal life conditions, but the comparisons that were made and the general mechanisms observed remain valid. In the same way, a modification of the characteristics of the garments (donors) was considered, after many transfers. However, a control made with pullover 7 (the most used) showed that it behaved in an identical manner after the first and after the hundredth contact. Finally, the absence of alteration of the collecting surface as a consequence of the numerous bands of tape that were applied was controlled by carrying out a succession of transfers with pullovers 4 and 9 on the cleaned passenger seat. The results confirm observations published elsewhere [6,10]; when a person sits in a car, fibres will be transferred from the garment worn to the seat, often in large numbers (more than a hundred to several thousands), even if the donor and the recipient have a smooth texture. All the parameters studied (donor, recipient, driving time and driver) have a significant influence on the transfer but the influence of the donor on the transfer is the greatest. Neither the composition of the garment (as a generic class) nor the

texture can explain by themselves the observed differences. Other characteristics (manufacturing process, weaving mode, etc.) certainly have an importance and combine with the other attributes. The driving time increases the number of fibres but this rise is not as important as was first thought. Indeed, the action of sitting down and getting up causes many more fibres to be transferred than the simple act of driving, and determines on its own the order of magnitude of the transfer. In sitting down and getting up, the contact is quite rough and provokes violent friction between the seat and the driver's garment, while in driving the transfer is much less dynamic and is due to a succession of much smoother rubbings. Pounds and Smalldon 1975 [3] proposed three general mechanisms for the transfer: first, transfer of loose fragments already on the surface of the fabric; secondly, loose fibres being pulled out of the fabric by friction; and thirdly, transfer of fibre fragments produced by the contact itself. We suggest the hypothesis that driving involves mainly the second mechanism, while the action of sitting down or getting up induces all three mechanisms. Our experiments suggest that there is a time limit after which the number of transferred fibres becomes constant. This does not mean that no fibres are transferred after this limit but rather that an equilibrium is reached between the fibres that transfer from the garment to the seat and those that return from the seat to the garment. Time also influences the distribution of fibres on the seat. If the person does not drive or only for a short time, the majority of the fibres are transferred near the left shoulder area, i.e., at the centre of all the swivelling movements (entry, fastening the seat-belt, getting out, etc.) in left-hand drive cars. Then, as the person drives longer, movements on the right hand side of the seat (changing gear, handbrake, etc.) will increase with the effect of a more even distribution of fibres over the seat. In either case, transfer will be most efficient where the friction is greatest and where the seat protrudes. Thus, the central area will contain less fibres than the peripheral areas because the middle of the back has a tendency to stay relatively immobile during driving and the structure of the seat emphasises the contact with the sides and the higher part of the arms. By combining our results with those of Robertson and Lim [lo] who divided the seat in a different way (from top to bottom), we can see, as intuitively expected, that the areas where we should find the highest number of fibres is at the top of the peripheral areas, which corresponds to the level of the shoulder blades. The influence of the recipient is significant but mostly low. Excluding leather seat as an extreme case, the transfer of fibres from a given garment is generally constant whichever seat is used (fabric, felt, hairy). Even if this result needs confirmation, it could be explained by the fact that the manufacturers of car seat fabrics submit their products to well Science & Justice 1996; 36(3): 143-151

C ROUX, J CHABLE and P MARGOT

defined conditions that give them transfer characteristics that are relatively constant. In the same way, the effect of the driver is low and the differences displayed are seldom significant, so that fibre transfer remains constant whoever wears the garment. Also, the number of transferred fibres is in no way correlated to the weight of the driver. This result, surprising at first, could be explained by a combination of one or other of the following phenomena: heavier persons are usually more corpulent and the local pressures associated with the weights remain approximately constant; these pressures are distributed over the saturation limit; the influence of the pressure on the number of transferred fibres is low when the recipient is soft. As all the results obtained during this study seem to show that the donor material is in most cases the determining factor, a good descriptor of this parameter should enable a simplified prediction for routine cases. Unfortunately, such a descriptor is difficult to find. As a matter of fact, neither the texture, nor the composition of a garment is sufficient to predict if a fabric is going to be a good or a bad donor. The sheddability test used in this study is easy and seems to give quite reliable results for a first estimation, but Coxon et al. [12] have warned that an adhesive tape lifts the superficial fibres, while a transfer is rather the result of friction. This test does not take into account the interaction that really exists between the donor and the recipient, and may overestimate the real sheddability of a garment especially if the garment in question is smooth. Other questions remain to be studied. For example, the different variables which have an effect on the number of transferred fibres were deliberately restricted. Likewise, other fabric donors and other seat recipients could have been used, but only the most common categories were tested in this study. Finally, the problems relative to fibre persistence will be considered for further research.

Conclusion In our experiments, it seems that the donor garment is the determining factor. An increase in driving time increases, in a logarithmic fashion, the number of transferred fibres, which are also more uniformly distributed on the seat. The influence of the recipient is generally low, and excluding leather seats, the number of transferred fibres from a given garment is constant whatever the seat and is not correlated to the weight of the driver.

Science & Justice 1996; 36(3): 143-151

These results have implications for routine investigations. It is always useful to search for fibres on car seats even if the suspected contact is brief andor involves smooth surfaces. A prediction based on experience can provide a useful first assessment but the only accurate way to estimate the number of fibres likely to be transferred is from experiments on the actual materials involved in the case. Further research is needed to find a realistic assessment of fibre shedding.

Acknowledgments The authors would like to thank all the persons (family, friends and colleagues) who provided cars, garments or participated to these experiments as drivers. They also gratefully acknowledge Mr Eric Lock, Eric Sapin and Eric Diirst, Institut de Police Scientifique et de Criminologie, University of Lausanne, for their valuable assistance. References 1. Locard E. L'enquCte criminelle et les mithodes scientifiques. Paris: Flammarion, 1920: 1392. 2. Pounds CA and Smalldon KW. The transfer of fibres between clothing materials during simulated contacts and their persistence during wear - Part 1: Fibre transference. Journal of the Forensic Science Society 1975; 15: 17-27. 3. Pounds CA and Smalldon KW. The transfer of fibres between clothing materials during simulated contacts and their persistence during wear - Part 3: A preliminary investigation of the mechanisms involved. Journal of the Forensic Science Society 1975; 15: 197-207. 4. Kidd CBM and Robertson J. The transfer of textile fibres during simulated contact. Journal of the Forensic Science Society 1982; 22: 301-308. 5. Kriston L. ijber den Beweiswert der Textilmikrospuren. Archiv fiir Kriminologie 1984; 173: 109-1 15. 6. Grieve MC, Dunlop J and Haddock PS. Transfer experiments with acrylic fibres. Forensic Science International 1989; 40: 267-277. 7. Evett IW. The theory of interpreting scientific transfer evidence. Forensic Science Progress 1990; 4: 141-179. 8. Stoney, DA. Transfer Evidence. In: Aitken CGG and Stoney DA, ed. The Use of Statistics in Forensic Science. Chichester: Ellis Horwood, 1991: 107-138. 9. Grieve MC. Fibres and forensic science - new ideas, developments, and techniques. Forensic Science Review 1994; 6: 60-79. 10. Robertson J and Lim M. Fibre transfer and persistence onto car seats and seatbelts. Ninth Australian International Forensic Symposium, Melbourne, Australia, February 1986. 11. Grieve MC and Garger EF. An improved method for the rapid and accurate scanning of fibres on tape. Journal of Forensic Sciences 1981; 26: 560-564. 12. Coxon A, Grieve MC. and Dunlop J. A method of assessing the fibre shedding potential of fabrics. Journal of the Forensic Science Society 1992; 32: 151-158.