A number of observers have noted the presence of bright structures near the cores of the chromospheric rosettes when observed in the far wings of the Hα line (eg Hα ±7/8 Å). Dunn and Zirker observed these bright structures with the highest possible resolution using the Sacramento Peak vacuum solar telescope. They find that these bright regions exhibit a very intricate fine structure which can be followed out much further into the Hα line wing (eg Hα + 2 Å) and even into the continuum. They called this fine structure ‘solar filigree’, the name referring mainly to the collective appearance of the fine structure elements. The elements themselves appear as dot-like structures and frequently also as small wiggly structures called ‘crinkles’. The properties of the filigree structure are summarized as follows: (i)Size: Measured diameter of the crinkles and dots equals 0.25, 0.40 and 0.60″ at Hα + 2 Å, Hα ± 7/8 Å and Hα ±5/8 Å respectively. The telescope resolution equals 0.22″ so that at Hα + 2 Å the structure is extremely small. The drawings in Figure 1 show typical sizes of the crinkles and network patterns in the filigree.(ii)Contrast: Filigree is enhanced in the blue wing of the Hα line. Measured contrast, uncorrected for seeing, equals 5–10%.(iii)Relation to the Granulation: The filigree structures tend to lie between the granules. This is, however, not a strict rule. It seems that in the course of their lifetime the granules move the filigree structures around with velocities of about 1.5 km s-1. Some of the crinkles also seem to wash out temporarily until compressed again by a new granule. The detailed structure of the filigree, therefore, changes significantly over times comparable to the granule lifetime. The overall structure is, however, preserved over much longer periods of time. The granulation pattern when observed in the continuum well outside the Hα line appears very peculiar in that it has substantially decreased in contrast. It appears ‘soft’ similar to granulation washed-out by seeing. This abnormal granulation can be traced over long times (> 30 min) and coincides in location to the filigree location. It is, therefore, definitely real.(iv)Relation to the spicules: The filigree structure falls near the center of the Hα chromospheric rosettes. These rosettes consist of dark elongated mottles which should probably be identified with spicules. There is, therefore, at least a coarse relation between the occurrence of spicules and the filigree. There is no clear evidence that variations in the filigree pattern are related to the generation of spicules. Some spicules seem to originate from the spaces between the crinkles. Too few, however, to conclude a definite relation.(v)Relation to the magnetic field: Beckers studied the filigree with the Universal Birefringent Filter in the magnesium b1 and b2 lines. It is very well visible in the far wing of the lines (eg. b1 ±0.8 Å). When traced into the line core the structures increase somewhat in size, as they do in Hα, and form structures similar to, and perhaps identical with, the so-called photospheric network. In the magnetically sensitive b2 line one sees a one-to-one correspondence between these network structures and the magnetic field so that, at least in the layers seen near the core of the b2 line, there is a one-to-one correspondence between the filigree structures and the enhancements in the magnetic field. Simon and Zirker (Solar Physics, submitted for publication) using a spectrograph also found that the filigree occurs in regions of enhanced magnetic field. However, in contrast to the filter observations, they found the magnetic field regions to be much more diffuse (2–3″) so that there is not a one-to-one spatial correspondence between filigree and magnetic field structure.