PKS 1814-637 is a compact steep spectrum radiosource (= -0.86) identified with a galaxy with a Seyfert 2 nucleus (Thom- pson et al., 1990). Unfortunately there is a galactic star (Tadhunter et al., 1993) which we locate at 17 SE from the nucleus accounting for about half of the total light of the combined image which makes accurate photometry and spectroscopy rather difficult. Danziger & Goss (1979) obtained a redshift of z= 0.0627, Thompson et al. (1990) z=0.065 and Tadhunter et al. (1993) 0.063. We have remeasured the spectrum of Thompson et al., kindly provided by Dr. S.G. Djorgovski and obtained from the [O III] lines z=0.0646 and from the H, [N II] and [S II] lines z=0.0642. We adopt z=0.064, but an error of 0.001 cannot be excluded.
A 15 min I image was obtained with the NTT on May 3, 1992, while with EFOSC three 10 min images were obtained in I and in B on July 27, 1995 and a one min image in I on the same day. Seeing conditions were far from ideal (16 at the NTT and 18-22 at EFOSC). Moreover in the 10 min EFOSC I images, the star near the centre is overexposed.
The subtraction of a stellar image does not give really satisfactory results, especially with the EFOSC images which have a pixel size of 0607/pix (the pixel size of the NTT image, 0129/pix, is more favorable).
We obtain a (B-I) colour of 1.4 for the star 17 SE from the nucleus.
The colour of the galaxy itself could only be determined with some confidence in a 49x85 box centered 14 W and 2 N from the nucleus. In that box, the influence of the star is negligible. Combining all the B and I images, we obtain B-I=2.15. For a gE, we would have expected 2.46.The radio galaxy is therefore too blue by 0.3 in B-I.
We have measured on the NTT I image the total intensity in a box 23 x 14 (42x26 kpc) centered on the nucleus after subtraction of the star 17 SE of the nucleus and found I=14.96. With B-I=2.15 this gives M=-22.3 if the B0-V0 colour is 0.2 less than for a gE.