•^m •^K^•e 291 samples were collected by retrieving those which fell out naturally during brushing and combing and only those which had an intact root were selected. Starting at the root end a 40 mm long segment was clamped between the cross heads of an Instron tensile tester and, at ambient room temperature and relative humidity, extended at various rates of extension until fracture occurred. Fractures were obtained in this manner from root to tip of the various fibres and all were mounted for examination in the SEM. Five main types of fracture were encountered and these are illustrated in Figs 1-5. Type 1 (Fig. 1) was found at the root end of all the hairs but was also observed towards the tips of the less weathered hair samples. It con- sists of a clean transverse fracture and appears almost as if cut with a knife. Also, the cuticle has split circumferentially about the transverse fracture through the cortex. Type 2 (Fig. 2) was also found near the root ends of the hairs. It consists mainly of a transverse fracture with the cortex stepped and with some disturbance of the cuticle behind the point of fracture, either in the form of a longitudinal split back from the main fracture or a narrow circumferential split some distance from the point of primary fracture. In type 2 (Fig. 3) part of the primary fracture is transverse but the remainder tails off with segments of cortex pulled out. Damage to the cuticle behind the point of fracture is more severe, usually in the form of lifting of the cuticle as well as more longitudinal and circumferential splitting than in type 2. Types 4 and 5 (Figs 4 and 5 respectively) are typical of fractures occurring close to the tip of extensively weathered hair. In type 4, although the hair may still retain its cuticle, there is no clean transverse fracture of the type encountered in types 1-3. Instead the fracture is ragged with the cortex separating into fibrillar elements (presumably individual cortical cells and macro fibrils) and there are nearly always one or more longitudinal splits back along the fibre from the point of primary fracture. In type 3 there is extensive fibrillation of the cortex resembling that encountered in tri- chorrhexis nodosa. There is a general progression in the type of fracture 1-5 proceeding from root to tip of the various hairs, with 2, 3, 4 and 5 being reached more quickly for badly weathered hairs than for hairs which have not been so extensively weathered. In the case of hairs which have not received appreciable weathering, types 4 and 5 are not encountered even at the tips of the hairs. The classification of fracture types has served well as a guide to the extent of degradation of the hair and correlates satisfactorily with other methods we have used for assessing the extent of hair weathering. By deter- mining the type of fracture at a given distance from the scalp and scoring

292 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS 1-5, some measure of quantification amongst the hair of the six women could be obtained. Although this is a useful method for assessing certain aspects of hair weathering it is evident from our examination of natural fractures that the circumstances leading to the fracture of hair on the head is more complex than simple tensile fracture of single hairs. It is already well known that one of the most damaging forms of grooming is the combing out of hair tangles and particularly those produced in backcombing (teasing). This aspect was therefore considered in more detail using the SEM. EXAMINATION OF TANGLED HAIR (STATIC STUDIES IN THE SEM) In this experiment a lock of hair was lightly backcombed (teased) and then partially combed out to the point where the usual tangling occurred. The tangled lock was mounted in the SEM. At low magnification a mass of fibres looped round each other was seen (Fig. 6). At high magnification it was possible to see where the hair cuticle of individual fibres had lifted and had been stripped off (Fig. 7). Where hairs passed over one another or had become twisted round each other, grooves had been formed in the cuticle surface (Fig. 8) and in some cases loose cuticle debris was observed. In such a complex situation it was difficult to assess the causes of these effects with- out watching the whole process as it was taking place. At this stage therefore we turned to an examination of the comb-out operation actually taking place within the SEM, the whole experiment being recorded on video tape. DYNAMIC COMBING EXPERIMENTS IN THE SEM Before giving details of these combing experiments it is necessary to elaborate on the conditions that must be satisfied for the successful use of the SEM. Hair fibres are electrically non-conducting and tend to accumulate an unstable electrostatic charge under the influence of the electron beam in the SEM which leads to instability of the video image. Usually this problem is overcome by providing the specimen with a thin surface film of metal evaporated under vacuum. Naturally such a metal film is undesirable in our present studies. Some authors have used antistatic coatings for eliminating electrostatic charging (9-11) but these are also undesirable in that they affect