AMERICAN JOURNAL OF PHYSICAI, ANTHROPOLOGY 79281-288 (1989) Developmental Changes in the Facial Soft Tissues P.H. BURKE AND C.A. HUGHES-LAWSON Department o f Child Dental Health, University of Sheffield, School of Clinical Dentistry, Sheffield. S10 2SZ. England KEY WORDS Craniofacial growth, Cephalometry, Somatic growth patterns, Neural patterns ABSTRACT Short-base stereophotogrammetry was used to study differential growth and development of the soft tissues of the face. Thirteen facial parameters were measured a t ages 9, 11, 13, 15, and 16 years on 170 facial contour maps selected from a mixed longitudinal study of 26 boys and 26 girls. Each parameter was measured three-dimensionally, and its developmental progress a t the earlier stages was expressed as a percentage of its value at 16 years of age. Standing height development was assessed in the same way. Three parameters that measured soft tissues surrounding the eyes grew little but were very advanced in their development, following a “neural” pattern. The remaining facial parameters grew more but were less advanced, and standing height was least advanced. There appeared to be three separate patterns of development, “neural,” “facial,” and “skeletal.” Girls were, in general, smaller than boys, but their development was more advanced when measured as a percentage of size at 16 years compared with boys. Craniofacial growth and development presents complex patterns that have been studied in the main by cephalometry, a method introduced by Broadbent in 1931 and followed up with a serial report in 1937. At about the same time, Zeller (1939)published a contour map of a face a s a n example of shortbase stereophotogrammetry. The accurate photography needed in cartography, which uses overlapping aerial photographs to produce three-dimensional contour maps of terrain, could be applied to the face and was used by Thalmaan-Degen (1944), Bjorn et al. (1954), Haga et al. (1964), and Berkowitz and Cuzzi (1977) to measure facial change in three dimensions. The complexity and expense of the plotting machinery discouraged its wider use. However, a simpler system was evolved (Beard and Burke, 1967), and its accuracy was investigated (Burke and Beard, 1967; Burke, 1971). This system was used to study facial growth. Serial study suggested the presence of a n adolescent growth spurt in the soft tissues of the face (Burke and Beard, 1979). Some of the records from this study were used for the present investigation. Previous knowledge about development of facial soft tissues has either been obtained by direct measurement, e.g., Davenport (1939), 6 1989 ALAN R. LISS, INC Meredith and Higley (1951),Meredith (1960), and many others, or was derived from the soft tissue midline profiles on serial cephalometric radiographs, e.g., Subtelny (1959), Posen (1967), Roos (1974), and Bishara et al. (1985). These studies revealed late facial development in the increasing convexity of the midline soft tissue profile when the nose tip was included in its measurement. Subtelny (1959)correlated the soft tissue changes with the underlying skeleton and showed that changes in the soft tissues do not reflect directly the skeletal changes. None of these studies included parameters of facial width. The aim of the original study was to investigate any relationship between growth of the soft tissues of the face and adolescent somatic growth as measured by standing height. Facial maps (Fig. 1)were plotted to allow retrospective study of any aspect of facial soft tissue growth found to be of interest at a later date. The cost of the maps became prohibitive, and the present system measures “x,y,z” coordinates of the landmarks (Burke, 1984). To the initial group of 13 nonoverlapping parameters were added eight further paramReceived June 2. 1987; revision accepted September 30, 1988. 282 P.H. BURKE AND C.A. HUGHES-LAWSON Fig. 1. Contour map of the face with a contourinterval of 2 mm posed with the Frankfort plane coincidental with the plane joining the central rays of the stereometric camera. 4 eters, to extend the range of measurement. From this group of 21 parameters all oblique parameters were excluded, leaving 13parameters, all vertical (or near vertical) or horizontal (Fig. 2), to simplify analysis of the findings for the changes of facial growth and development. MATERIALS AND METHODS The data for this study were derived from a selection of 170 facial maps selected from a mixed longitudinal study of 26 boys and 26 girls between the ages of 9 and 16 years (Burke and Beard, 1979). The technique for recording the stereo pair of facial photographs has been reported (Burke and Beard, 1967). The subject was supine, and the face was posed by means of -2- 6 / T P-TI l3 12 1 \&/ 5-- 7 4 X Fig. 2. A selection of three-dimensional linear facial parameters based on soft tissue anatomical and olottinaiandmarks. 283 FACIAL SOFT TISSUE CHANGES TABLE 1. Means and standard deviations of 13 facial parameters in mm and standing height in cm for boys at 9, 1 1 , 13, 15, and 16 years of age Parameter 1 2 3 4 5 6 7 8 9 10 11 12 13 Standing height 9 (n = 8) 30.3 f 2.4 27.6 f 1.7 27.1 f 1.7 33.1 f 1.3 45.8 f 3.1 30.4 f 1.6 28.8 f 3.2 16.1 f 3.8 16.9 f 3.4 62.2 f 4.8 62.8 i 4.7 85.4 f 8.6 82.2 f 4.0 134.6 f 6.4 11 (n = 14) Ape - in vears " 13 (n = 21) 15 (n = 17) 16 (n = 16) 30.5 k 2.3 28.4 k 2.7 28.8 f 4.1 34.4 f 2.4 47.2 f 2.3 31.9 2.3 29.9 i 3.5 15.6 f 3.0 18.4 f 3.8 63.7 f 5.0 63.7 f 4.4 85.8 f 6.2 84.0 f 5.9 140.2 i 6.6 30.8 f 2.2 29.5 f 3.5 28.4 f 2.6 38.4 f 2.4 48.1 f 1.2 33.4 3.3 32.3 f 4.4 16.4 f 3.9 19.7 f 3.8 66.9 f 5.5 66.4 f 5.6 92.0 i 9.2 86.6 f 5.8 155.6 f 8.7 31.8 f 2.1 29.5 k 2.0 29.2 f 1.9 39.5 f 2.7 51.0 f 4.8 37.1 f 3.0 33.0 f 3.0 17.9 f 4.8 21.8 f 4.2 70.9 f 4.0 70.6 f 3.8 98.0 k 7.8 88.4 i 5.4 169.2 f 6.7 31.6 i 2.8 29.6 i 2.4 29.2 1.9 39.6 f 2.9 51.1 i 3.5 37.2 i 2.3 33.6 f 3.4 15.7 2.8 21.8 f 2.9 71.2 i 2.8 71.0 i 3.3 97.8 f 4.4 88.0 f 4.2 172.4 f 5.4 + ear rods and a n optical pointer, locked at the same level as the ear rods, to focus on to left orbitale, which is marked on the face. In this way, the Frankfort plane was aligned with the plane joining the axes of the two optical systems in the camera. The maps were measured a t ages 9,11,13, 15, and 16 years, and the numbers of children in the different age groups are given in Tables 1 and 2. The numbers of maps measured for boys and girls were 76 and 94, respectively. The maps were life size with a contour interval of 2 mm (Fig. 1)and included the eyes, nares, and mouth, extending in depth 50 mm from the tip of nose, which was given zero value. All children were healthy like-sexed twins of European descent. The value for 16 years for each parameter was * * * given 100%status, and the percentage value for the various parameters at the other ages measured their relative growth and development. The selection of facial parameters was controlled by the availability of soft tissue anatomical landmarks, which can be recognised with reliability when the overlapping facial stereo pair of photographs are projected to form a three-dimensional image of the face. It is important to realise that linear parameters defined by two landmarks within the plot are not subject to a posing error. A posing displacement results in the contours being recorded differently, but they compensate exactly to record the same distance threedimensionally. To reduce location error of tip of nose this landmark was plotted four times TABLE 2. Means and standard deviations of 13 facial parameters in mm and standing height in em for girls at 9, 1 1 , 13,15, and 16 years of age Parameter 9 (n = 6) 11 (n = 20) Age in years 13 (n = 20) 15 (n = 24) 28.2 f 2.3 25.4 i 3.9 26.8 i 0.5 32.1 f 2.0 43.0 f 3.1 30.7 3.1 27.4 f 2.4 13.7 f 2.5 18.1 f 2.5 62.1 1.3 62.0 f 2.0 16.7 f 2.5 74.8 f 3.2 133.1 f 6.4 29.7 f 2.0 27.3 f 2.1 27.1 f 1.9 33.1 f 2.8 45.5 f 3.3 31.8 f 2.6 29.0 k 2.4 14.0 & 2.2 18.2 2.5 65.1 f 2.7 64.6 3.0 82.1 f 5.5 81.0 f 3.6 143.7 f 5.8 30.2 f 2.7 27.4 f 2.6 27.8 f 3.1 34.7 f 2.6 47.7 f 3.5 34.0 i 2.6 30.1 i 1.7 14.2 f 2.2 18.3 f 2.8 66.0 f 2.3 65.6 f 2.3 84.8 i 5.6 83.0 t 3.8 154.0 f 6.3 31.0 2.8 26.6 f 2.3 27.7 f 2.5 35.7 2.4 47.9 2.4 34.7 i 2.8 31.1 f 2.1 13.9 f 2.1 18.8 f 3.2 66.8 k 2.4 66.5 f 2.4 87.8 k 6.0 83.0 f 4.0 161.0 f 4.4 * 10 11 12 13 Standing height * + * + + * 16 (n = 24) 30.0 i 2.8 27.5 i 2.1 27.4 f 1.6 35.5 i 2.8 47.7 f 2.6 33.9 f 2.5 30.9 i 2.3 14.4 f 1.6 18.9 i 3.2 67.0 3.5 66.9 f 3.5 88.0 f 4.2 83.8 i 3.2 163.1 i 4.2 * 284 P.H. BURKE AND C.A. HUGHES-LAWSON and mean values used for coordinates. In view of the small facial growth velocities, the absence of posing error was important. The parameters used in this study are illustrated in Figure 2 and are defined below: 1. Right internal canthus to left internal canthus 2. Right internal canthus to right external canthus 3. Left internal canthus to left external canthus 4. Right external ala of nose to left external ala of nose 5. Right angle of mouth to left angle of mouth 6. Skin overlying midpoint of intercanthal line to tip of nose (vide infra) 7. Tip of nose to midpoint of vermilion border of upper lip 8. Midpoint of vermilion border of upper lip to corresponding point on lower lip 9. Midpoint of vermilion border of lower lip to soft tissue pogonion 10. Right external canthus to right angle of mouth 11. Left external canthus to left angle of mouth 12. Skin overlying midpoint of intercanthal line (as in parameter 6) to soft tissue pogonion 13. Right external canthus to left external canthus Soft tissue pogonion was a landmark difficult to locate in subjects with little chin prominence. In such children each individual series of maps was examined to achieve a consistency of location relative to the pattern of contour lines. Plotting definitions were required for three landmarks: 1. Tip of nose: the centre of a circle of diameter 4 mm that rests on the summit plateau of nose tip. 2. Ala of nose: the most lateral contour line defining the ala with a distance of 1 mm or more on the map to the adjacent lateral contour contained within the cheek. 3. Mid canthal point: this rests on the skin overlying the midpoint of the intercanthal line measured on a plane parallel to the Frankfort plane. ence for the landmarks to give a threedimensional expression. The values were age-corrected using the transformation ratios on growth charts (Tanner et al., 1966), and means and standard deviations were calculated for each parameter for boys and girls for the ages 9, 11, 13, 15, and 16 years. Accuracy of the method The original analysis of variance (Burke, 1971)was based on the same face being posed and photographed on three separate occasions, and two plots being constructed from each facial stereo pair. The standard deviation was 0.69 mm for repeated observations of 6 sets of 13 linear parameters (not subject to posing error); 0.65 mm was due to plotting error. The maps were originally measured manually with a dial calipgauge. In this study, the maps were measured on a digitisor on which 200 repeated observations gave standard deviations of 0.17 mm in “x” and 0.206 mm in “y.” The “z” coordinate of the landmark was read from the map and fed into the computer. If the landmark was not on or immediately near a contour line, the value for the “z” coordinate was obtained by interpolation and, depending upon its distance from the two adjacent contour lines, was estimated to a n accuracy of 0.5 mm. Reproducibility of this value was fairly consistent, but the plotting variance in “z” still remained, and the “z” coordinate was subject to posing error. To investigate the variance in “z” further, the “z” values for all 11 landmarks were abstracted from the six maps of the original analysis of variance. The standard deviation for the ‘‘z” coordinate for 24 repeated measurements of the eight landmarks from the three stereo pairs was 0.279 mm; thus, the variance of the “z” values was slightly larger than those for the “x” and “y” values. The only comparable values for three-dimensional measurements are those of Savara and Singh (1966),who measured bony growth of the face three-dimensionally on lateral and posteroanterior cephalometric radiographs. All his standard deviations of linear parameters were greater than 1.0 mm. RESULTS The values of the parameters were derived Means and standard deviations of facial by measuring the distance between the land- parameters in millimetres and standing marks on the maps using a n x,y,z digitiser height in centimetres are reported for the varand incorporating the contour height differ- ious ages in Tables 1 (boys) and 2 (girls). FACIAL SOFT TISSUE CHANGES 285 TABLE 3. Relative growth expressed as a percentage of its own size at 16 years of age for the same 13 parameters at ages 9, 11,13, and 15 years for boys and girls Age in years Girls Boys Parameter 1 2 3 4 5 6 7 8 9 10 11 12 13 Standing height 9 11 13 15 9 11 13 15 96 93 93 84 90 82 86 100 78 87 88 87 93 78 97 96 97 89 92 86 89 99 84 89 90 88 95 81 98 100 97 97 94 99 96 100 99 94 96 94 98 90 100 100 100 100 100 100 99 100 100 100 99 100 100 98 94 92 98 90 90 91 89 95 96 93 93 a7 89 81 99 99 99 93 95 94 94 97 96 97 97 93 97 88 100 100 100 98 100 100 97 100 96 98 98 96 99 94 100 100 100 100 100 100 100 100 99 100 100 100 99 99 These values are expressed as percentages of their sizes at 16 years in Table 3 for boys and girls. In general, girls are seen to be smaller than boys, but, in terms of proportional size, are more advanced at comparable ages, Most facial parameters grow appreciably between 9 and 16 years. The exceptions are 1, 2, 3, and 13 (eye measurements), 8 (mouth height), and 9 (chin height) in girls. Parameters 1,2,3, and 13may be considered as “neural” since the eyes are an extension of the central nervous system and therefore follow the neural pattern (Scammon, 1930).Parameter 8 measuring mouth height shows little change in either sex, and this may be related to the need, early in life, for a sphincter for suckling and subsequent change of function to a slit for access for food conveyed by finger and thumb. The other parameter that grows little in girls, 9 (chin height), agrees with the recent findings of Buschang et al. (1986). All other simple facial parameters in this study grow well, as is demonstrated by the means and the proportionate development when expressed as percentages (Table 3). The composite parameters 12 (facial height) and 13 (facial width at eye level or biocular width) reflect the changes of their individual components. Facial height shows considerably greater growth changes, whereas facial width (at eye level) shows less growth (except for the increment between 9 and 11 years for girls). These changes reflect the typical lengthening of the face in the older child. Mean percentage sizes for the two groups of nonoverlapping parameters neural (1-3) and facial (4-7, 9-11-but excluding 8) are reported in Table 4. Interestingly, the corresponding values for standing height are less advanced a t all levels of development for both sexes. The differentiation of neural, facial, and somatic growth patterns at 9, 11, and 13 years is graphed in Figures 3 and 4 for boys and girls, respectively. In boys, the developmental paths of the facial and neural parameters are completely separate until age 13 years when they begin to merge. In girls, this separation is less clearly defined at 9 years (parameter 2 is based on a small sample size). Nevertheless, by 11years it is complete. By 13 years a convergence of all parameters occurs. The percentage values of standing height at 9, 11, and 13 years in both boys and girls are at lower levels than the facial values. TABLE 4. Means of ‘heural,” “facial,” and ‘<skeletal” parameters at ages 9, 11, 13, and 15 years for boys and girls Age (veard Boys 9 11 13 15 Girls 9 11 13 15 Mean percentage of size at 16 years Neural Facial Skeletal 94 97 98 100 85 88 96 100 78 81 90 98 95 99 100 100 92 95 98 100 81 88 ~~ 94 99 286 P.H. BURKE AND C.A. HUGHES-LAWSON BOYS GIRLS IW 95 90 ,‘ I 70 a 85 80 I I 75’ a,’ S H 15 I I I 9 in 11 I 12 , 13 1 1 14 15 7 16 J , , , 9 10 11 I 12 13 14 15 16 hie (Years1 A x lyearrl Fig. 3. Relative growth at various ages expressed as a percentage of size at 16 years, of neural, facial, and skeIetal parameters for boys. Fig. 4. Relative growth at various ages expressed as a percentage of size at 16 years, of neural, faciaI, and skeIetal parameters for girls. DISCUSSION Their conclusions were as follows: “A distinct facial pattern of growth is established. In terms of the proportion of final size achieved during childhood, it is below the cranial pattern but above the general skeletal one.” A similar study based on serial lateral cephalometric radiographs was carried out by Buschang et al. (1983) on 26 boys and 26 girls between the ages of 4 and 16 years. This study revealed similar findings, namely three separate patterns of growth and development for neural, facial, and skeletal organs. Further, they proposed a continuous developmental gradient running from the cranium to the upper and lower face similar to Tanner’s (1962, 1978) suggestion for the trunk and limbs. The findings in this study support the concept of a developmental gradient between neural and general somatic growth with soft tissue facial growth following a n intermediate path. Somatic growth is characterised by its sustained nature in the second decade and a n adolescent spurt (Tanner, 1962). The fact that the face appears not to follow this pattern exactly may explain the difficulties experienced in demonstrating a facial adolescent growth spurt on serial cephalometric records (Broadbent et al., 1975). Scammon (1930) was among the first workers to draw attention to the developmental differences between the general skeletal and the neural longitudinal growth curves. While most of the rapid neural growth takes place in the 1st decade of life, skeletal growth follows a sigmoid curve until the pubertal growth spurt, after which it decelerates and essentially ceases as the end of the second decade approaches. Since the face grows well into the 2nd decade (Broadbent et al., 1975)it has been presumed that the face follows the latter pattern of development. Evidence for an adolescent growth spurt has been presented by Nanda (1955), Bjork (1964), and Savara and Singh (1966).These studies were based on lateral and frontal cephalometric radiographs, and Savara and Singh evaluated three-dimensional parameters. Baughan et al. (1979) studied serial lateral cephalometric records for 50 French-Canadian girls between the ages of 6 and 15 years and arrived at different conclusions. Mean size changes for cranial and facial parameters were expressed as a proportion of the size at 15 years to measure development. Standing height was treated in the same way. FACIAL SOFT TISSUE CHANGES This study also reveals some coordination between developmental pattern of parameters concerned with anatomical entities on the face. For example, parameters 4 and 6, which measure nasal width and nasal height, both follow a relatively delayed path of development compared with many other facial parameters. In contrast, mouth width (5) and height (8) are advanced. This is true in both sexes, but at the age of 9 years cheek height (10 and 11) is relatively advanced in girls compared to boys. The growth patterns of cheek height (10 and 11) are similar to those of facial height in the midline (12), while facial width growth slows down. Thus, the differential velocity of growth of the vertical and horizontal parameters produces a n overall change in facial shape. These findings suggest that growth and development of the soft tissues of the face are organ orientated, and priority of development varies between different organs and in time. Particular examples are the eyes and within the face, the nose, and the mouth. This agrees with the recent findings of Bailey (1986) who measured genetic variation of shape of the mandible in CXB strain of mice. This study was based on a two-dimensional multivariate analysis of MLDs (minimal linear distances) between anatomical points and revealed that genetic control is organorientated, e.g., for the lower incisor area. This also agrees with the findings of Kraus et al. (1959) for bony craniofacial characteristics, as shown in cephalometric radiographs in six sets of triplets. The findings in this study provide some support for Buschang et al.’s concept of a craniofacial developmental gradient between the cranium and middle third of face, but the continuation of the gradient to the lower third of face is supported only in boys by the developmental pattern of the one parameter (9), which is the only parameter to be contained entirely within this part of the face. The results do, however, support the concept that genetic control is organ orientated, and further analysis may help to clarify the position. ACKNOWLEDGMENTS The authors wish to thank Mr. L.F.H. Beard, formerly Director, Department of Medical Illustration, University of Cambridge, for his sustained help over many years with stereophotogrammetry. They wish 287 to acknowledge, with gratitude, funding from the United Cambridge Teaching Hospitals Board, University of Sheffield, and the Medical Research Council of the U.K. They are also grateful for the commercial plotting services provided by Fairey Surveys Ltd. and Clyde Surveys Ltd. Finally, they wish to thank the children in the study for their patience. LITERATURE CITED Bailey DW (1986) Genes that affect morphogenesis of the murine mandible. J. Hered. 77:17-25. Baughan B, Demirjian A, Levesque GY, and LapalmeChaput L (1979)The pattern of facial growth before and during puberty, as shown by French-Canadian girls. Ann. Hum. Biol. 6:59-76. Berkowitz S, and Cuzzi J (1977) Biostereometric analysis of surgically corrected abnormal faces. Am. J. Orthod. 72526-538. Beard, LFH, and Burke PH (1967) The evolution of a system of stereophotogrammetry for the study of facial morphology. Med. Biol. Illustr. 17:20-25. Bishara S, Hessian T, and Peterson L (1985)Longitudinal soft tissue profile changes. Am. J. Orthod. 88:209-221. Bjork A (1964) Sutural growth of the upper face studied by the implant method. Trans. Eur. Orthod. Sac. 49-65. Bjorn H, Lundquist C, and Hjelstrom P (1954) A photogrammetric method of measuring the volume of facial swelling. J . Dent. Res. 35:295-308. Broadbent H (1931)A new x-ray technique and its application to orthodontics. Angle Orthod. 1:l-45. Broadbent H (1937) The face of the normal child. Angle Orthod. 7:183-208. Broadbent BH, Sr., Broadbent BJ, Jr., and Golden WH (1975) Bolton Standards of Dentofacial Developmental Growth. St. Louis: C.V. Moshy Co. Burke PH (1971)Stereophotogrammetricmeasurement of normal facial asymmetry in children. Hum. Biol. 45:536-548. Burke PH (1984) Four dimensional facial change. Br. J. Orthod. 11:170-184. Burke PH, and Beard LFH (1967) Stereophotogrammetry of the face. Am. J . Orthod. 53:769-782. Burke PH, and Beard LFH (1979) Growth of the soft tissues of the face in adolescence. Br. Dent. J. 146: 239-246. Buschang P, Baume R, and Nass G (1983) A craniofacial growth maturity gradient for males and females between 4 & 16 years of age. Am. J. Phys. Anthropol. 61:373-381. Buschang PH, Tanguay R, Demirjian A, la Palmer L, and Goldstein H (1986) Sexual dimorphism in mandibular growth in French-Canadian children 6-10 years of age. Am. J. Phys. Anthropol. 71:33-37. Davenport CB (1939)Postnatal development of the human outer nose. Proc. Am. Philos. Sac. 80:175-356. Haga M, Ukiya M, Koshihara Y, and Ota Y (1964) Stereophotogrammetric study of the face. Bull. Tokyo Dent. Call. 5:l-24. Kraus B, Wise N, and Frei R (1959) Heredity and the craniofacial complex. Am. J . Orthod. 45:172-217. Meredith HV (1960) Changes in the form of the head and face during childhood. Growth 24:215-264. Meredith HV, and Higley LB (1951)Relationship between dental arch width and width of head. Am. J. Orthod. 37:193-204. 288 P.H. BURKE AND C.A. HUGHES-LAWSON Nanda RS (1955) The rates of growth of several components measured from serial cephalometric radiographs. Am. J. Orthod. 53:746-756. Posen J (1967) Longitudinal study of growth of nose. Am. J. Orthod. 53:746-756. Roos N (1974) The Soft Tissue Profile: A Cephalometric Study Dissertation. Stockholm: Karolinska Institute. Savara B, and Singh J (1966) Norms of size and annual increments of seven anatomic measures of maxillae in boys from three to sixteen years of age. Hum. Biol. 38:105-120. Scammon RE (1930) In J A Harris, CM Jackson, DG Patterson, and RE Scammon (eds.): The Measurement of Man. Minneapolis: Univ. of Minnesota Press. Subtelny D (1959) Longitudinal study of soft tissuefacial structure in relation to underlying skeletal structures. Am. J. Orthod. 45:481-507. Tanner JM (1962) Growth at Adolescence. Oxford: Blackwell Scientific Publications. Tanner JM (1978)Foetus Into Man. Cambridge,MA: Harvard Univ. Press. Tanner JM, Whitehouse RH, and Takaishi M (1966) Standards from birth to maturity for height and weight. Arch. Dis. Child. 41:454-471, 613-635. Thalman-Degen P (1944) Die Stereophotogrammetrie ein diagnostiche hilfsmittel in der Kieferorthopadie. Doctoral Dissertation. Univ. Zurich. Zeller M (1939) Textbook of Stereophotogrammetrie. Zurich, translated into English by Miskin AE, 1952. London: H.K. Lewis & Co. Ltd.