2 Observations and reductions

2.1 B and R band observations

Prime focus CCD images were obtained with the Calar Alto 3.5 m telescope during a photometric night, the details are given in Table 1. After debiasing and flatfielding in the usual manner, I constructed an image which contains only the underlying smooth light distribution of the unresolved fainter stars as described by Hopp & Schulte-Ladbeck (1991). This "smooth mask'' was used to derive the structural parameters by applying the ellipticity fit of Bender & Möllenhoff (1987). Then, the "smooth masks'' were subtracted from the original B and R images which yielded frames with the resolved stars only. To these frames, DAOPHOT in its MIDAS version was applied, again following Hopp & Schulte-Ladbeck (1991). The frames of both colors were searched independently by DAOPHOT and only those objects were accepted, which were found in both frames. Figure 1 shows the B image of UGC 685.

2.2 H${\alpha}$ observations

The Calar Alto 2.2 m telescope and its CCD camera were used to obtain H${\alpha}$- and R-images of UGC 685. The night was of poor photometric quality, thus only a preliminary calibration through the R band observations was applied. Details are given in Table 1. As above, the usual CCD calibration frames were obtained and applied and the individual exposures in each filter were added. The R image was flux-scaled to the H${\alpha}$-image with the aid of several medium-bright stars and then subtracted, yielding a continuum free H${\alpha}$ frame of UGC 685 (for more technical details see Bomans et al. 1997). Figure 2 shows the H${\alpha}$ image of UGC 685.

2.3 J band observations

J-band observations were taken with the Calar Alto 3.5 m telescope and its IR camera MAGIC during a photometric night. The set-up details can be found in Table 1. After every five object frames, nearby night sky data were taken at 4 different positions around UGC 685. The sets of object frames were slightly shifted against each other. The individual 10 s. exposures were corrected for bias, dark current, night sky pattern and fixed-pattern noise with the usual calibration data. Then they were transformed into a common reference frame and finally added.

As the on-line facility did not show any signal of UGC 685, no K data were obtained. For the same reason, centering was not perfect and the eastern outskirt of UGC 685 was therefore not mapped within the tiny field of view available with this set-up.

  

Table 1: Details of the observations presented in this paper

  

Table 2: Overall properties of UGC 685 from the literature and as derived in this paper. Absolute magnitudes are corrected for galactic reddening with the given extinction value from Burstein & Heiles (1982), distance depend values are converted with the new distance estimate of 5.5 Mpc

  

Figure 1: CCD 3.5 m telescope B image of UGC 685. North top, East left. The horizontal side length is 346$^{\prime\prime}$. The letters identify those very bright stars where residuals are still visible in Fig. 2

  

Figure 2: CCD 2.2 m telescope image in H${\alpha}$ of UGC 685 after continuum subtraction. North top, East left. The side length is 160$^{\prime\prime}$. The residuals seen at large distances from UGC 685 are from saturated stars which are identified in Fig. 1 by letters

  

Figure 3: Top: The surface brightness profiles according to the ellipticity fits in the red (upper) and blue (lower) 3.5 m CCD frames. Middle: The color gradient as calculated from the curves in the upper graph. The straight line gives the color for the total light. Bottom: The radial ellipticity variation derived from the blue (squares) and red (triangles) frames (all 3.5 m telescope CCD data)