Introduction

Since the discovery of low metallicity gas in H II regions of Blue Compact Galaxies (BCGs) by Searle & Sargent (1972), a debate lasting for decades on possible existence of local truly young galaxies is still continuing. While great majority of BCGs have metallicities Z in the range of (1/10 - 1/3) Z $_\odot$ , or respectively 12 + log(O/H) in the range 7.92 - 8.42 very few galaxies have O-abundance as low as 12 + log(O/H) = 7.1 - 7.6, consistent with the expected oxygen enrichment produced in one star formation (SF) episode. Izotov and Thuan ([1999]) from the analysis of variations of C and N abundances have shown, that namely the BCGs with 12 + log(O/H) $\leq$ 7.6 are the best candidates to the galaxies with the first SF episode. To clear up the true evolution status of these extremely metal-poor BCGs is a serious challenge for observational astrophysics. However, independently on whether their current SF burst is the first or second one, properties of these local well resolved galaxies best approximate those of primeval low-mass galaxies at high redshifts. Therefore systematic studies of group properties of this small sample are very important from cosmological point of view.

Kunth & Östlin ([1999]) (KÖ), in their review of the problems related to the most metal-poor galaxies, summarized published up-to-now data and presented the compilative list of these galaxies. In course of studies of new large samples of BCGs in a broader context, we attempt to find as well new very metal-poor galaxies (e.g. Kniazev et al. [1998], [2000a], [2000b]). In parallel we check old measurements for several BCGs, claimed to have very low metallicities. In particular, we reobserved two BCGs presented in the KÖ list - UM 133 and UM 382. Reticon spectrophotometry for these two BCGs is originally from Terlevich et al. ([1991]) and Masegosa et al. ([1994]).

In this letter we present a new spectrophotometry with the SAO RAS 6m telescope for UM 133 and UM 382. For UM 133 we give also the H $\alpha$ velocity curve and B-band photometry. For UM 283 with its published emission line intensities we revised its claimed low O abundance.

**Figure 1:** From top-left to bottom-right: a) UM 133 brightness profile of H $\alpha$ in relative intensities along the slit (lower profile is scaled by factor 1/8 to show the contrast of bright features) and the velocity curve. The error bars for the latter correspond to $\pm 1\sigma$ . The solid line represents the linear fit; b) UM 133 surface brightness profile in B-band with bright SW spot subtracted, and its exponential disc fitting; c) B-image of UM 133 with the position of long slit superimposed. The brightest feature on the H $\alpha$ intensity profile corresponds to the bright spot ``a'' at the SW edge of the galaxy; d) 1-D spectrum of the spot ``a'' in UM 133 with dispersion 4.6 Å/pixel; the lower one is scaled by factor of 1/10 to show relative intensities of strong lines. In the inserted frame we show the broad WR blue bump and indicate 3 characteristic WR-lines. e) DSS image of UM 382 with the long slit position superimposed; f) 1-D spectrum of UM 382 with dispersion 2.4 Å/pixel; the lower one is scaled by factor of 1/10 to show relative intensities of strong lines.
$\begin{figure} \vspace{-3.0cm} \psfig{figure=Figure1.ps,angle=0,width=19cm}\vspace{-3cm} \end{figure}$

**Table 1:** Line intensities of the studied galaxies
	UM 133 $^{\bf b}$		UM 382		UM 283 $^{\bf c}$
$\lambda_{0}$ (Å) Ion	F( $\lambda$ )/F(H $\beta$ )	I( $\lambda$ )/I(H $\beta$ )	F( $\lambda$ )/F(H $\beta$ )	I( $\lambda$ )/I(H $\beta$ )	F( $\lambda$ )/F(H $\beta$ )	I( $\lambda$ )/I(H $\beta$ )
3727 [O II] ...............	1.047 $\pm$ 0.019	1.093 $\pm$ 0.023	0.396 $\pm$ 0.026	0.639 $\pm$ 0.046	5.24	5.74
3868 [Ne III] ............	0.290 $\pm$ 0.023	0.301 $\pm$ 0.024	0.328 $\pm$ 0.031	0.491 $\pm$ 0.050	--	--
4101 H $\delta$ ...................	0.245 $\pm$ 0.014	0.269 $\pm$ 0.022	0.155 $\pm$ 0.026	0.285 $\pm$ 0.064	--	--
4340 H $\gamma$ ...................	0.450 $\pm$ 0.010	0.472 $\pm$ 0.018	0.330 $\pm$ 0.018	0.452 $\pm$ 0.039	0.39	0.41
4363 [O III] ..............	0.083 $\pm$ 0.008	0.084 $\pm$ 0.008	0.114 $\pm$ 0.015	0.134 $\pm$ 0.018	--	--
4861 H $\beta$ ..................	1.000 $\pm$ 0.013	1.000 $\pm$ 0.017	1.000 $\pm$ 0.023	1.000 $\pm$ 0.037	1.00	1.00
4959 [O III] ..............	1.286 $\pm$ 0.017	1.263 $\pm$ 0.017	2.152 $\pm$ 0.041	1.942 $\pm$ 0.040	1.15	1.14
5007 [O III] ..............	3.743 $\pm$ 0.039	3.668 $\pm$ 0.039	6.554 $\pm$ 0.125	5.801 $\pm$ 0.119	3.24	3.21
6312 [S III] ...............	0.017 $\pm$ 0.001	0.016 $\pm$ 0.001	--	--	--	--
6563 H $\alpha$ ..................	2.990 $\pm$ 0.060	2.782 $\pm$ 0.062	5.342 $\pm$ 0.094	2.816 $\pm$ 0.059	3.15	2.86
6584 [N II] ...............	0.034 $\pm$ 0.001	0.032 $\pm$ 0.001	--	--	0.25	0.23
6717 [S II] ................	0.112 $\pm$ 0.004	0.103 $\pm$ 0.004	--	--	0.42	0.38
6731 [S II] ................	0.076 $\pm$ 0.003	0.071 $\pm$ 0.002	--	--	0.39	0.35

C(H $\beta$ ) dex	0.08 $\pm$ 0.03		0.77 $\pm$ 0.02		--
EW(abs) Å	2.70 $\pm$ 2.14		5.30 $\pm$ 2.24		--
F(H $\beta$ ) $^{\bf a}$	133 $\pm$ 2		79 $\pm$ 2		136
EW(H $\beta$ ) Å	193 $\pm$ 2		82 $\pm$ 1		122

$^{\bf a}$ in units of 10^-16 ergs s^-1cm^-2; $^{\bf b}$ the intensities of the lines for red part are taken from spectrum with
1.2 Å/pixel; $^{\bf c}$ all data are taken from Gallego et al. ([1996]).