A re-evaluation of the premaxillary bone in humans
- First Online:
- Cite this article as:
- Barteczko, K. & Jacob, M. Anat Embryol (2004) 207: 417. doi:10.1007/s00429-003-0366-x
The discovery of the premaxillary bone (os incisivum, os intermaxillare or premaxilla) in humans has been attributed to Goethe, and it has also been named os Goethei. However, Broussonet (1779) and Vicq d’Azyr (1780) came to the same result with different methods. The first anatomists described this medial part of the upper jaw as a separate bone in the vertebrate skull, and, as we know, Coiter (1573) was the first to present an illustration of the sutura incisiva in the human. This fact, and furthermore its development from three parts:—(1) the alveolar part with the facial process, (2) the palatine process, and (3) the processus Stenonianus—can no longer be found in modern textbooks of developmental biology. At the end of the nineteenth and in the early twentieth century a vehement discussion focused on the number and position of its ossification centers and its sutures. Therefore, it is hard to believe that the elaborate work of the old embryologists is ignored and that the existence of a premaxillary bone in humans is even denied by many authors. Therefore this re-evaluation was done to demonstrate the early development of the premaxillary bone using the reconstructions of Felber (1919), Jarmer (1922) and data from our own observations on SEM micrographs and serial sections from 16 mm embryo to 68 mm fetus. Ossification of a separate premaxilla was first observed in a 16 mm embryo. We agree with Jarmer (1922), Peter (1924), and Shepherd and McCarthy (1955) that it develops from three anlagen, which are, however, not fully separated. The predominant sutura incisiva (rudimentarily seen on the facial side in a prematurely born child) and a shorter sutura intraincisiva argue in this sense. The later growth of this bone and its processes establish an important structure in the middle of the facial skull. Its architecture fits well with the functional test of others. We also focused on the relation of the developing premaxilla to the forming nasal septum moving from ventral to dorsal and the intercalation of the vomer. Thus the premaxilla acts as a stabilizing element within the facial skeleton comparable with the keystone of a Roman arch. Furthermore, the significance of the premaxillary anlage for the closure of the palatine was documented by a synopsis made from a stage 16, 10.2 mm GL embryo to a 49 mm GL fetus. Finally the growth of the premaxilla is closely related to the development of the human face. Abnormal growth may be correlated to characteristic malformations such as protrusion, closed bite and prognathism. Concerning the relation of the premaxillary bone to cleft lip and palate we agree with others that the position of the clefts is not always identical with the incisive suture. This is proved by the double anlagen of an upper–outer incisor in a 55 mm fetus and an adult.
KeywordsArchitecture of craniofacial regionHistoryHuman embryosPremaxillary bone developmentSutures
In the development of the human craniofacial morphology, a special part of the palate maxilla nose region—the premaxillary bone (os incisivum, os intermaxillare, os Goethei, Zwischenkiefer)—plays an essential role. According to developmental biologist, the anlagen of this bone belong to the primary palate (Peter 1911, 1913, 1924; Hochstetter 1949, 1955). It contains the alveolar ridge of the four upper incisors, and was observed in animals by Galen (cited by Vacher et al. 2001a) due to sutures delimiting the secondary palate, maxillary, nasal and frontal bones; however, Vesalius 1538 (in Gysel 1993) neglected a special bone for man. The discovery of this bone in humans was related to Goethe (1784, see for references Bräuning-Oktavio 1956). Yet other earlier reports do exist.
In actual textbooks of embryology, if mentioned, the premaxillary bone is illustrated in relation to the facial and palatine development in a lingual view, showing an at first bipartite, then fused trapezoid and finally a wedge-shaped horizontal body, which is rostrally intercalated into the two palatine halves of the maxillary. Today the frontal, as well as the nasal processes (crista nasalis or processus Stenonianus after Jarmer 1922, prevomerine center after Fawcett 1911, or os praevomerale after Wegner in Peter 1924) with the foramen incisivum are ignored.
Likewise illustrations in osteological atlases show only the nasal and lingual side with the corresponding suture. There is no demonstration of the frontal process which fuses with the maxillary bone facially and which punctually attaches itself to the frontal bone.
Not only the number of ossification centers of maxillary and/or premaxillary bone is controversial (Kölliker Th. 1882, 1888; Felber 1910; Jarmer 1922; Peter 1924; Chase 1942; Woo 1949; Shephard and McCarthy 1955; Kraus and Decker 1960; Vacher et al. 1999, 2001a and others), but even the existence of a separate anlage has been questioned (Fawcett 1911; Eschler 1966; Wood et al. 1967; Vacher et al. 1999, 2001a, 2001b).
We therefore designed this article: (1) to discuss that, Goethe cannot be considered as the first discoverer of the “Zwischenkiefer” in humans; (2) to demonstrate, that the premaxilla develops in man as a separate three-dimensional structure composed of three different parts. In this point we especially refer to two nearly forgotten investigators, Felber (1919) and Jarmer (1922) with their plastic reconstructions, illustrating origin and development of the “Zwischenkiefer” element, and their own serial sections of embryos from 16 mm to 68 mm; (3) to show that the premaxilla exert an important function as a vertical stabilization element in the architecture of the facial skeleton, including all four upper incisors; (4) to consider the premaxillary bone development in relation to the movement of the nasal septum during the fusion of the secondary palate; (5) to analyze data from the literature and own embryos concerning the association of the premaxillary bone to the primary facial processes and thus to the location of alveolar clefts.
History regarding the discovery of the intermaxillary bone in humans
The discussion about the existence of a separate intermaxillary bone in humans seems to be nearly as old as the history of comparative anatomy. The first known description of this bone can be dated back to Galen (131–200) (see for details Franz 1933; Ashley-Montagu 1936) in animals. Vesalius (1543, in Gysel 1993), however, denied the existence of such an independent bone in humans. Since then, the intermaxillary bone (“Zwischenkiefer”) became a central place in the discussion of dissimilarity or homology between men and other vertebrates. Thus, in the concept of Camper 1778 (cited by Bräuning-Oktavio 1956) the division of maxilla “entre la dent incisoire et la canine” is considered to be a main difference between ape and man. In this context Goethe’s discovery of the “Zwischenkiefer” at the human skull has to be valuated. Its historical circumstances are precisely investigated and described by Bräuning-Oktavio (1956). According to Goethe, accident and reflection had guided him in his research on several skulls of animals and men. It is not only the anatomical discovery that has made the work of Goethe so famous and important, but it was Goethe’s idea of a common type—harmonia naturae (Typus, Bauplan)—of the vertebrate body that contradicted against a special position of men (Goerttler 1950).
To Herder, 27th March 1784, “Ich habe gefunden—weder Gold noch Silber, aber was mir eine unsägliche Freude macht—das os intermaxillare am Menschen! Ich verglich mit Lodern Menschen- und Thierschädel, kam auf die Spur und siehe da ist es.” And to Mrs v. Stein the same day, “...ich habe eine anatomische Entdeckung gemacht die wichtig und schön ist.”
In the original paper, Loder was also helpful with the right terminology and the translation into Latin. Many years later (1819 and 1831) official “de luxe” editions were published.
According to Lubosch (1931), the early eighteenth century was the time of the first flowering of comparative anatomy and it was especially Vicq d’Azyr who—like Goethe—looked for similarities among the species. It was also Vicq d’Azyr who discovered (see for references Bräuning-Oktavio 1956) the intermaxillary bone, which he called os maxillaire antérieur, in the human fetus and presented his “Mémoires” at the Academy of Sciences in Paris 1780. Thus, it is the latter to whom the priority in this research field is due, even if both Goethe and Vicq d’Azyr came independently and with different methods to similar results. Beyond that, in 1779, Broussonet (cited by Bräuning-Oktavio 1956) read his first “Mémoire” about the teeth and the intermaxillary bone, yet an official edition of his work appeared later in 1789. Nevertheless Vicq d’Azyr did not cite him and required the priority. But Goethe surely learned about the work of Vicq d’Azyr from Loder, but never cited Vicq d’Azyr. Kohlbrugge 1913 (cited from Franz 1933) in his critical studies of Goethe’s scientific work, reproached him for holding back information.
In the nearly forgotten work of Volcher Coiter (1573, 1573/1955, Fig. 1c) (see also his biography in Herrlinger 1952) an illustration of the sutura incisiva can be found, although he only wrote “Superioris maxillae ossa ossibus adultorum figure, et numero, et situ respondent”.
Although Goethe was not the first to discover the premaxillary bone as again stated recently by Neddermeyer (2002). The great philosophical and ideological importance of this topic can be estimated by Rousseau’s (1858) treatise “de la non-existence de l’os intermaxillaire” arguing in the sense of Camper (see for references Franz 1933; Bräuning-Oktavio 1956). Thus the debate about this bone did not finish: Lawrence (1844) surgeon and comparative anatomist denied its existence in man, yet Leidy (1849) demonstrated the existence of intermaxillary bone in a 9 to 10 week-old human embryo. Likewise, Callender (1869) who investigated the ossification of the bones of the human face found that in a fetus of 4.3 inches “the intermaxilla completely formed and that it may be traced back as a distinct bone. In a fetus 9 inches long, either intermaxillary bone is in great part fused with its corresponding upper maxilla.” Interestingly he writes in his introduction that it was its purpose “to supply information regarding some few points with which we are yet imperfectly acquainted, such as the growth of the maxilla, and the formation and eventual obliteration of the intermaxillary bones.”
The development of the premaxilla
In the nineteenth and early twentieth century, the discussion about parts of the intermaxillary bone and its sutures was settled again with great vehemence and especially the controversy between Albrecht and Kölliker Th. is well reported (Inouye 1912).
The main discussion now focused on the number of ossification centers. The most important authors, in our view, to this topic are summarized as follows (based on data of Kräutler 1966):
Two centers of intermaxilla: Authenrieth 1879 (see for references Felber 1919; Inouye 1912), Albrecht 1879, 1884, v Bardeleben 1879, Biondi 1888, v. Meyer 1884, Thourén 1926 (cited by Kräutler 1966), Wallesch-Gladzinski 1995, Warynski 1888.
According to Le Double (1906) in Bardeen (1910), some authors describe six or more centers (maxilla and premaxilla taken together). In the discussion of multiple ossification centers, Albrecht (1879) introduced a sutura intraincisiva, and Biondi (1888), a sutura intraalveolaris that are supposed to exist beside the generally accepted sutura incisiva and sutura palatina media.
A second heavily debated question was from which facial processes the intermaxillary bone derives. Kölliker Th. (1882, 1888) named only the tissue of the medial nasal process as the source. According to Albrecht (1879), Warynski (1888) and v. Meyer (1884) the bone derives from the mesenchyme of the medial nasal and the maxillary process. Inouye (1912) and Jarmer (1922) even traced it back to the medial and lateral nasal and the maxillary process.
These opinions are closely related with the question of synonymy of the primary palate and intermaxillary bone on one hand, and the location of clefts in lip and palate on the other. Inouye (1912) could solve this problem by demonstrating, that the development of bones does not coincide with the primary development of the face. As a consequence the epithelial-fusion lines are not identical with the incisive fissure (Lisson and Kjaer 1997).
A three-dimensional figuration of the developing bone seems to be a good tool for a better understanding of its relationships. The elaborate graphic reconstructions of Jarmer (1922) on young human embryos and fetuses, and Felber’s graphic (1919) of 5 month-old fetuses are worth represented in a new modus.
(a) The 25 mm (eighth week) embryo exhibits the bony anlage of the alveolar part of the premaxilla already fused with the maxillary anlagen. The primary palate deriving from the medial nasal process is dedicated. Its border to the lateral facial processes is located more medially than the boundary between alveolar parts of premaxillary and maxillary anlage. We can conclude that the medial part of the maxillary process delivers material to the premaxilla.
(b) The 40 mm (ninth week) embryo shows the separate anlage of the palatine process that completes in the horizontal view the premaxillary bone.
(c) Greater than 40 mm: the palatine processes of the paired premaxillary bone are fused with the body of the premaxilla. Dorsal, a sutura incisiva reminds on the boundary between each palatine process of the premaxilla and the maxilla. Two little ovoid bones arise in the median plane, considered as anlage of processus Stenonianus, or spina and crista nasalis anterior.
At stage 22 (26 mm GL) the paired primary palate grows medially and fuses in the midline (Fig. 4b, d). The palate processes of the maxilla approach each other. As the demarcation between primary and secondary palate the sutura incisiva of the soft tissue can be considered (Fig. 4a). Its deeper prolongation is not identical with the bony sutures.
According to Hinrichsen (1991) the epithelial remnants indicating the fusion line of facial processes has disappeared first, in the frontal part of a stage 18 embryo (15 mm GL). The premaxillary ossification center appears shortly afterwards when fusion is completed at stage 18, 16 mm (see Fig. 6a). An exact correspondence between facial processes and bony anlagen do not exist, though the main part of the premaxilla arises within the medial nasal swelling.
From our results and studies of the literature we conclude that in the human the existence of the premaxilla can be regarded without doubt, the arguments are summarized in the following text.
The appearance of the premaxillary bone was reported by Mall (1906) as early as in a 16 mm human embryo; this is in line with our own observation. According to Fawcett (1911) ossification begins in a 17 mm embryo. Chase (1942) found the premaxilla in all embryos longer than 21.5 mm at first in 19 mm as a separate bone with a tendency of the premaxilla to appear later then the maxilla. He also stated that both bones had fused before the 25 mm stage. Concerning the palatine process he summarized that it normally grows directly from the body but may be separated in same cases. This also holds true for the Stenonian process.
Shepherd and McCarthy (1955), investigating fifteen human embryos, came generally to similar results like Jarmer (1922) who found a separate palatine process that soon fuses with the body. In the latter case the palatine process was not fully isolated, but thin-bony bridges connected it with the premaxillary corpus. Therefore the decision whether three isolated centers exist is a problem which is not conclusively resolved. Kölliker Th. (1882) vehemently argued for only one ossification center, although he only presented a few young embryos treated in caustic lye. Drawings from the bone visualized in such a manner are not very detailed. Histological sections from these filigree-like structures are absent and 30 to 50 μm-thick serial sections are shown from a 68 mm embryo only.
The authors who favor three ossification centers of premaxilla (Jarmer 1922, Peter 1924, Shepherd and McCarthy 1955) did not find totally isolated elements but rather fine trabeculae connections which show their belonging to the premaxilla. This is also the case in our material. Further evidence pointing toward separate anlagen is the persistence of the sutures. Beside the sutura incisiva between the palatine part of premaxilla and maxilla, a sutura intraincisiva separates the processus Stenonianus from the palatine process. Thus, we conclude that ossification of the premaxilla starts from three anlagen rather than to spread out from one center. However, we are of the same opinion as Peter (1924): “der morphologische Wert eines Skelettstücks wird also nicht durch die Zahl seiner Ossificationszentren gegeben, und damit verliert die Frage nach der Zahl der Anlagen der Intermaxillare ganz bedeutend an Wichtigkeit.”
While the number of more or less separated parts of the premaxilla is discussed, in recent times even the opinion that no separate ossification center of the premaxilla exists has found recent support. Wood et al. (1967) studied histological sections from embryos of the Carnegie Collection and were unable to find separate centers. Vacher et al. (1999, 2001a, 2001b), likewise, favor a single zone of ossification, in their opinion, it only persists a transverse mesenchymal septum as a “souvenir of an ancestral premaxilla.” In 1921, Macklin failed to note the very recognizable premaxilla with the sutura incisiva to the adjacent part of the maxilla in his famous reconstruction of the skull of a human fetus which was 43 mm at its greatest length. Müller and O’Rahilly (1979), in their detailed study on the skull of a stage 23 embryo do not mention the premaxilla. However, Kadanoff et al. (1970) found in 11.1% of palates in adults a sutura incisiva or remnants of it.
Despite its short individual existence, the existence of an individual premaxillary bone might be proved by an isolated agenesis of this bone. Fischel (1905) described the skull of a woman lacking the incisive teeth as well as crista nasalis and spina nasalis, a part of the upper jaw that clearly belongs to the premaxilla. Here the canines are the most medial teeth.
Another argument is the evolutionary aspect, Maurielle and Bar (1999) suggested a longer independence of some parts of the premaxilla in very young Neandertal children at the nasal aspect traces of the premaxillary suture. Yet, Neandertals are considered not to be a common ancestor to modern humans (Krings et al. 1997).
Finally, the phylogenetic aspect has to be taken into account as discussed by many authors since the time of Goethe as mentioned above.
The architecture of the premaxilla
The significance of this part of the face is well documented by Benninghoff’s trajectories of the skull, showing a strong line of stress named the frontonasal pillar (Benninghoff and Goerttler 1961) or nasal pillar (Toldt 1914, cited by Witzel and Preuschoft 2002). This was recently confirmed by the finite element analysis (Witzel 2002, Witzel and Preuschoft 2002). According to these authors “the lateral walls of the nasal cavity are the most highly stressed pillars that connect the upper jaw with the braincase, particular at both sides of the apertura piriformis.” Similarly, Lautrou (2002) shows a “centromaxillaire” bordering the capsule nasal.
The premaxilla development in relation to the movement of the nasal septum
A reconstruction block with a dorso-ventral view (Fig. 12b) proofs the keel-shaped nasal septum not to come from the base of the skull, but arises ventrally and grows backwards synchronously with the palatine shelves closure. Thus, the point of contact between nasal septum and palatine is an essential hallmark.
The final relation of the premaxilla to the septum nasi is reported in detail by Mosher (1909) and Klaff (1956): at birth, two structures of the premaxilla, the anterior nasal spine and the premaxillary wings become component parts of the nasal septum. The premaxillary wings project obliquely upward and fuse with the tip of the vomer at 15 years of age. According to Mosher (1909) “the tip of the vomer rest in the gutter of the premaxillary wings, and the tip of the premaxillary wings rests in the gutter of the nasal spines, like the arrangement of the sections of the old-fashioned V-shaped wooden drain.”
Synopsis of palatine closure
Thus palatal closure is not only the fusion of the palatal shelves but a soldering of three structures as already described. The palatine processes and the nasal septum start to fuse at stage 23. Complete fusion seems to be incomplete until the 12th week (Yoon et al. 2000). SEM images of Waterman (1974) demonstrated alterations of surface epithelium associated with cell death during the fusion process in humans.
In our last stage (49 mm), bone forms within the secondary palate. Only isolated epithelial nests remind the palatal fusion superficial (Luke 1976). Structure and glycogen content in this epithelial seam was studied by Meller and Barton (1979).
Fusion and elevation of the palatal shelves, at least in mice, is combined with the accumulation of glycosaminoglycans (Knudsen et al. 1985). Especially, the expression of chondroitin sulfate proteoglycan on the apical surface of the seam epithelial is important in palatal shelf adhesion and is supposed to be regulated by TGF-b3 (Gato et al. 2002) while BMP and Shh signals seem to regulate the growth of the anterior region of palate (Zhang et al. 2002). Recently, the role of the transcription factor Tbx22 in palatogenesis was discussed (Bush et al. 2003; Herr et al. 2003). Expression is found in the inferior nasal septum and the palatal shelf before fusion.
The premaxilla in relation to the development of the human face
Disturbance of the facial outgrowth may cause protrusion, closed bite or prognathism. Since facial structures develop from the facial processes which consist of neural crest derived mesenchyme (Noden 1984, 1988), defects in the signaling interactions of epithelium and mesenchyme, as reviewed by Francis-West et al. (1998) and Francis-West et al. (2003) might cause such malformations and Msx-1 and Pax 6 seem to be involved in premaxillary development.
In the later fetal and postnatal period facial outgrowth is due to bone growth and the growth pattern of sutures (see for references Meikle 2002), perhaps under the influence of TGF-b (Adab et al. 2002).
Rudimentary premaxillary bones are generally found to be combined with other defects. In fetuses with holoprosencephaly, the premaxilla was often absent or reduced (Arnold et al. 1998b; Kjaer et al. 1997). Etiological factors affected with these malformations are supposed to act at an early embryological stage (Sedano 1970) like disturbance of the mesoderm anterior to the notochord (Kjaer et al. 1997). Data on experimentally-induced holoprosencephaly in the golden hamster even suggest that alterations in selected proteins in the prechordal mesoderm and adjacent floor plate, impair the adjacent neural plate and neural crest and, thus, induce these anomalies (Coventry et al. 1998). Similarly, the anticancer agent suramin was found to cause midfacial defects in chicken (Männer et al. 2003).
The relation of premaxillary development and lip/palate clefts
According to the morphogenetic classification of craniofacial anomalies by Pfeifer (1986), Gundlach and Pfeifer (1981), defects or aplasia of the premaxilla belong to the prosencephalic anomalies. In the area between the forehead and the lateral/back region of the head, which is a more passive zone during development, most malformations of the head are found with lip palate clefts being predominant. These are caused by multifactorial variations.
Mooney et al. (1991) suggest growth deficits from 14 to 21 weeks causing clefts rather than reduced mesenchymal ingrowth into the nasal fin at an earlier stage as proposed by Diewert and Shiota (1990) who found a reduced thickness. Thus the role of BMPs that regulate proliferation and patterning of facial mesenchyme has been studied by Ashique et al. (2002). Similarly, Tgf-betas and Msx2 are suggested as having a role in active postnatal facial growth (Adab et al. 2002). Furthermore, the role of epithelium in the pathogenesis of cleft lip and palates was pointed out by Millicovsky and Johnston (1981) and Millicovsky et al. (1982), although a combination of genetic and environmental factors may cause these defects.
Describing the existence, development and impact of the premaxillary bone during the facial development in humans, we have arrived at the following conclusions.
1. Goethe (1784) was not the first discoverer of this bone in humans. Prior to him, Coiter (1573), Broussonet (1779) and finally Vicq d’Azyr (1780) observed and/or described this structure in the human. The conclusive literature to this topic is discussed.
2. At the end of the nineteenth and early twentieth century, a dispute arose between the great anatomists concerning parts of the anlage of premaxilla and consequently the number of ossification centers, as well as the position of the incisive suture and the existence of other suturae e.g., sutura intraincisiva, sutura intraalveolaris. This is not the topic today, but again there are scientists who are against the original development of the intermaxillary bone. To demonstrate early stages of premaxillary development we used reconstructions of the nearly forgotten embryologists Felber 1919, and Jarmer 1922, and our own serial sections of embryos from 16 mm to 68 mm. We are convinced by these observations that an original premaxilla exists. Its ossification was first observed in a 16 mm embryo. We are favoring three pairs of anlagen which are not completely separate centers but connected by thin bony lamellae: (1) the alveolar part including the facial process, (2) the palatine part, and (3) the processus Stenonianus. Prevail sutura incisiva and a shorter sutura intraincisiva argue in this sense. Fusion of the premaxillary and maxillary bones is nearly complete in the fifth month. Parts of the palatine as well the facial part of the incisive suture may rarely persist. In a 42 cm CHL (Crown-Heel Length) prematurely born child, a short section of its facial part is still visible.
3. The premaxilla is considered to be very important stabilizing element of the facial skull. Together with the ossa nasales it encloses the apertura piriformis and punctually contacts the ossa frontales. It contains the upper incisors. Together with the maxillary bones, it fulfils an important supporting function in biting and mastication. A vertical bony pillar connecting the basis of the skull dorsocranially consists of the premaxilla (crista nasalis) and the intercalated vomer. The comparison with the keystone in a Roman arch, the removal of this would cause the collapse of the suggested arch itself.
4. The mesenchymal anlage of the premaxilla as primary palate plays an important role in the palatinal closure and therefore the separation of the oral cavity from the pharynx and the two nasal ducts. The palatinal anlagen of the premaxilla are the pioneers in the fusion process, followed by closure of the secondary palatine. Simultaneously, the nasal septum, intercalated caudally with the premaxilla, grows dorsad to contact the basis of the skull and thus separates the nasopharyngeal space. Interestingly, the growth of the nasal septum always delays the fusion of the palatine a little. In a synopsis this process of closure is documented in embryos from 10.2 mm to 49 mm GL.
5. The formation of the human face is characterized by the simultaneous extension of the head and the outgrowth of the facial skull. A measuring line between hypophysis and nasal pit compared in embryos from 10 mm to 39 mm GL helps to make this process visible. An abnormal outgrowth may cause protrusion, closed bite or prognathism.
6. It was not the aim of this paper to focus on the development of cleft lip and palate. Although we wish to make some comments: the position of the clefts is not always identical with that of the incisive suture, demarcating the bony premaxilla. An example is given by the double anlage of an upper outer incisor in a 55-mm fetus and an adult. Cleft lip and palate are supposed to have a multifactorial etiology. They may occur very early within the epithelial–mesenchymal anlage or later after the fusion of facial processes with possible duplication of the lateral–upper incisor
We particularly thank Dr. med. Sigurd Große-Oetringhaus (Dortmund) for reading the manuscript and Dr. med. Heinz Jürgen Jacob who kindly provided the SEM in Figure 4 c, d and e. We are grateful to Mrs. Annegrit Schlichting for technical assistance, Mrs. Antje Jaeger and Mrs. Marion Otto for expert photographic work.