In 20 specimens for longitudinal sections of the recti, we consistently observed a connective tissue band inserting into the limbus of the conjunctiva
9,10 (see Introduction) and a tendon of the medial rectus (MR) and lateral rectus (LR) muscles. This “limbus” was located far lateral or medial to the corneal limbus that has recently been noted as a site of the epithelial cell progenitors.
15,16 The connective tissue band for the MR was much thicker than the LR at all the stages examined. Type 1 collagen fibers in the connective tissue band, stained red in the present silver impregnation, were much less dense than those in the tendon. The anterior end of the muscle belly was located at the level of the lens until 15 weeks, but it moved posteriorly after 20 weeks. Muscle fibers in the MR and LR appeared not to be heterogeneous, and we found no morphologic demarcation between the orbital and global parts of the muscle in the longitudinal sections (
Figs. 1,
2).
Figure 1A shows the topographic anatomy of the anterior part of the MR at 15 weeks, and
Figure 1B shows that at 30 weeks. Although the eyeball and its surrounding structures grew in size more than twofold during this period (almost threefold from 12 to 30 weeks of gestation), the anterior part of the MR and LR showed the same combination of structures: a thin tendon slip from the muscle belly to the eyeball and a thick connective tissue band extending from the orbital surface of the rectus to the conjunctiva. The connective tissue band appeared to originate from the orbital surface of the medial or lateral recti. The attaching site of the band to the MR was at or in the anterior side of the equator of the eyeball, whereas the band attached to the LR in the posterior side of the equator. Thus, the lateral connective tissue band was longer than the medial band.
Figure 2 displays higher magnification views of
Figure 1. We noted a connection between the muscle endomysium (rich in type 4 collagen; indicated by black in silver impregnation) and the connective tissue band (type 1 collagen; indicated by red in silver impregnation;
Fig. 2B). Such fibrous connections increased with age. At the superior and inferior margins, the band was connected with a sheathlike structure for the inferior or superior rectus. Thus, the sheaths of the four rectus muscles, as a whole, appeared to provide a sleeve system surrounding the eyeball, as described by Koornneef.
17 Tenon's sheath or capsule was evident after 25 weeks at the global side of the rectus (
Fig. 1B): this “portal” morphology was the same as that in adults.
5,18
Cross-sections of the recti from three fetuses at 20 weeks clearly showed a site-dependent difference in morphology of the rectus sheaths. First, a sheath of the MR was not distinct in the posterior site (
Fig. 3A); second, the orbital-sided sheath suddenly increased in thickness (
Fig. 3B); third, the global-sided sheath gradually became unclear (
Fig. 3C); and fourth, only the orbital-sided sheath remained as a thick connective tissue band (
Fig. 3D). This sequence was the case with the LR: the connective tissue band suddenly became thick at attachment sites of the orbital-sided muscle bundles. According to our measurement using cross-sections, the MR and LR exhibited the maximum sectional area in the posterior side of the equator of the eye ball (MR, 0.47 mm
2 at mean; LR, 0.69 mm
2 at mean). However, at the anterior site in which the muscle bundles attached to the connective tissue band, both muscles were reduced in thickness (MR, 0.25 mm
2 at mean; LR, 0.55 mm
2 at mean). The orbital-sided muscle bundles attaching tightly to the connective tissue band reached 14% to 15% (or 18%–20%) in the cross-sectional area of the MR (or the LR). Silver impregnation exhibited abundant type 1 collagen fibers in the orbital-sided muscle bundles (
Fig. 3F). However, most of the orbital-sided muscle bundles of the MR and all the LR muscle bundles did not “insert” into the connective tissue band: the insertion was seen in only one or two bundles of the MR (1.5%–1.8% of the cross areas;
Figs. 3E–G). Moreover, the limited muscle insertions appeared to contain no or few type 1 collagen fibers for tensile stress (
Fig. 3G).
Taken together with results from both of the longitudinal sections and the cross-sections, the distinct connective tissue band to the conjunctiva was “originated from” the orbital surface of the MR and LR rather than the reverse (i.e., the muscles inserted into the connective tissue band).