Our observations in 1998 suggest a period of 4 $.\!\!^{\rm h}$ 6 $\pm$ 0 $.\!\!^{\rm h}$ 2 with an amplitude of 0.13 $\pm$ 0.01. Composite diagrams calculated with previously published periods between $\l$ $4\hbox{$.\!\!^{\rm h}$ }3-4\hbox{$.\!\!^{\rm h}$ }4$ have much larger scatter. The light-time corrected composite diagram is presented in Fig. 2. The zero phase is JD 2451162.3169.

Based on earlier data (see Table 3), a new model has been determined with amplitude method. The observed amplitudes vs. ecliptic longitudes with the fit are plotted in Fig. 3. The resulting triaxial ellipsoid has the following parameters: $a/b=1.15\pm 0.07$ , $b/c=1.05\pm0.05$ , while the spin vector's coordinates are $\lambda_{\rm p}=15/195~\pm~ 25^\circ$ , $\beta_{\rm p}=52~\pm~ 15^\circ$ , respectively. We could not reduce the observed amplitudes to zero solar phase, since the actual value of m parameter (e.g. Zappala et al. 1990) could not be estimated by the data sequence or asteroid classification. Also we have to note that a mixture of V and R amplitudes was used, thus the model should be considered as an approximate one. The O-C' model has also been determined (Fig. 4). For reducing the errors, lightcurves obtained between October, 1983 and February, 1984, were composed and one time of minimum was determined from this composite lightcurve. The resulting sidereal period is $P_{{\rm sid}}$ = 0 $.\!\!^{\rm d}$ 1964156 $\pm$ 0 $.\!\!^{\rm d}$ 0000001 with retrograde rotation.

$\begin{figure}\par ~~~ \leavevmode \psfig{figure=ds1752f02.eps,width=160mm}\end{figure}$

Figure 2: The composite R lightcurve of 683 (symbols: solid squares - December 14; open circles - December 16)

$\begin{figure}\par ~~~ \leavevmode \psfig{figure=ds1752f03.eps,width=160mm}\end{figure}$

Figure 3: The observed amplitudes vs. longitudes with the determined fit for 683 Lanzia

**Table 3:** Published photometries of 683 Lanzia
Date	$\lambda$	$\beta$	$\alpha$	A	$t_{{\rm min}}$	ref.
1979 03 19, 20	182	-27	9	0 $.\!\!^{\rm m}$ 12	43963.452	(1)
1982 12 16	49	9	11	0.14	45319.591	(2)
1983 10 12, 13	130	-9	18	0.15	45650.871	(2)
1983 11 15	137	-13	19	0.15	45650.871	(2)
1984 02 21	129	-23	10	0.16	45650.871	(2)
1987 10 19	16	23	7	0.12	47118.538	(2)
1998 12 14, 16	20	18	17	0.13	51162.275	p.p.

References: (1) - Carlsson & Lagerkvist (1981) (2) - Weidenschilling et al. (1990).

It was discovered by J. Palisa in Vienna, on October 21, 1911. To our knowledge, the only one photometry of 725 in the literature is that of Di Martino et al. (1994) carried out in 1985. They determined a sinodic period of 3 $.\!\!^{\rm h}$ 749 associated with a full variation of 0.3 mag. Our observations do not exclude that period, as they suggest a possible value around 4 hours. Unfortunately the data cover only 3 hours, thus we could not draw a firm conclusion. The observations were made under fairly unfavourable conditions, which is illustrated with the comp-check curve bearing a relatively high scatter (about $\pm$ 0.03 mag). It is presented together with the observed lightcurve in Fig. 5.

This asteroid was discovered by S. Belyavskij in Simeis, on April 2, 1916. Its earlier photometric observations were carried out in 1977, 1982 and 1993 (Tedesco 1979; Di Martino & Cacciatori 1984; De Angelis & Mottola 1995). The observed light variation in 1998 had an amplitude of 0.32 mag, while the period was 4 $.\!\!^{\rm h}$ 62 $\pm$ 0 $.\!\!^{\rm h}$ 01. This is in very good agreement with results by De Angelis & Mottola (1995), who found a period value of 4 $.\!\!^{\rm h}$ 613. The light time corrected composite diagram is presented in Fig. 6. The zero phase is at 2451160.5904. The lightcurve has remarkable asymmetries - the brighter maximum is rather sharp, its hump is exactly two times shorter than the other one. There are also small amplitude, short-period humps on the longer descending branch. These phenomena can be more or less identified in the previous measurements too. That is why we carried out a second observing run on January 24, 1999. We wanted to check the reality of these irregularities. The lightcurve revealed the same asymmetries as those of observed one month earlier (Fig. 7). This may refer to a shape with sharp asymmetries, e.g. something similar to a jagged tenpin.

$\begin{figure}\par ~~~ \leavevmode \psfig{figure=ds1752f06.eps,width=160mm}\end{figure}$

Figure 6: The composite R lightcurve of 852 (symbols: solid circles - December 12; dotted circles - December 14; solid squares - December 16)

**Table 4:** Published photometries of 852 Wladilena
Date	$\lambda$	$\beta$	$\alpha$	A	ref.
1977 02 14	139	31	10	1 $.\!\!^{\rm m}$ 12	(1)
1982 10 18	6	-10	10	0.37	(2)
1993 11 8, 10	33	-8	3	0.23	(3)
1998 12 12-16	163	23	19	0.32	p.p.
1999 01 24	170	19	14	0.27	p.p.

References: (1) - Tedesco (1979) (2) - Di Martino & Cacciatori (1984) (3) De Angelis & Mottola (1995).

$\begin{figure}\par ~~~ \leavevmode \psfig{figure=ds1752f08.eps,width=160mm}\end{figure}$

Figure 8: The observed amplitudes vs. longitudes with the determined fit for 852 Wladilena

This Earth-approaching asteroid was discovered by E. Hertzsprung in Johannesburg, on September 25, 1929. There are four photometric observations in the literature (Hahn et al. 1989; Velichko et al. 1990; Hoffmann & Geyer 1990; Chernova et al. 1995) and one radar measurement by Ostro et al. (1990). The previously determined periods scatter around 4 $.\!\!^{\rm h}$ 8, thus our resulting 4 $.\!\!^{\rm h}$ 80 $\pm$ 0.01 is in perfect agreement with earlier results. The amplitude changed significantly over a period of one month, as it was 0.77 mag and 0.92 mag in December, 1998 and January, 1999, respectively. The composite lightcurve is presented in Fig. 9, while the single lightcurve obtained in January is plotted in Fig. 10.

$\begin{figure}\par ~~~ \leavevmode \psfig{figure=ds1752f09.eps,width=160mm}\end{figure}$

Figure 9: The composite R lightcurve of 1627 (symbols: solid squares - December 14; dotted circles - December 15; crosses - December 16)

**Table 5:** Published photometries of 1627 Ivar
Date	$\lambda$	$\beta$	$\alpha$	A	$t_{{\rm min}}$	ref.
1985 06 13	317	29	48	0 $.\!\!^{\rm m}$ 35	46226.750	(1)
1985 08 31	15	-21	32	0.55	46258.703	(1)
1985 10 16	4	-23	20	0.63	46287.184	(1)
1989 05 01-23	203	25	20	1.0	47647.402	(2)
1989 06 15-23	201	21	51	1.12	47647.402	(2)
1989 07 14-19	213	14	60	1.45	47721.565	(2)
1990 05 11-14	204	25	24	1.08	48029.439	(3,4)
1998 12 14,16	76	-12	79	0.77	51162.295	p.p.
1999 01 26	87	-13	18	0.92	51201.171	p.p.

References: (1) - Hahn et al. (1989) (2) - Chernova et al. (1995) (3) - Velichko et al. (1990) (4) Hoffmann & Geyer (1990).

A new amplitude model has been determined after collecting all available data (Table 5). The observed amplitudes were reduced to zero solar phase. First of all, the m parameter was derived from our measurements. The observed amplitudes in December, 1998 and in January, 1998 were compared. As the longitudes differ by only 10 $^\circ$ , and the difference between the corresponding phases is quite high (13 $^\circ$ ), the amplitude change can be mostly associated with the phase change. The result is m=0.018. We have also corrected other amplitudes to zero solar phase and fitted the amplitude variations along the longitude. The corresponding parameters are: $a/b=2.0\pm 0.1$ , $b/c=1.09\pm 0.05$ , $\lambda_{\rm p}=145/325\pm8^\circ$ , $\beta_{\rm p}=34\pm6^\circ$ . The reduced amplitudes with the determined fit is presented in Fig. 11. The reliability of this model was tested by a direct comparison with radar images of Ostro et al. (1990). This is shown in Fig. 12, where we used Fig. 5 taken from Ostro et al. (1990) with kind permission of the first author. The similarity is evident.

The O-C' method was used to determine the sidereal period and the sense of the rotation. The results are $P_{{\rm sid}}$ = 0 $.\!\!^{\rm d}$ 1999154 $\pm$ 0 $.\!\!^{\rm d}$ 0000003, retrograde rotation with $\lambda_{\rm p}=143~\pm~8^\circ$ , $\beta_{\rm p}=-37~\pm~6^\circ$ pole coordinates. The agreement between the poles obtained by different methods is very good. The sidereal period agrees well with results of Lupishko et al. (1986) - 0 $.\!\!^{\rm d}$ 19991, prograde -, but the senses are in contradiction. The fitted O-C' diagram is presented in Fig. 13.

$\begin{figure}\par ~~~ \leavevmode \psfig{figure=ds1752f11.eps,width=160mm}\end{figure}$

Figure 11: The reduced amplitudes vs. longitudes with the determined fit for 1627 Ivar

$\begin{figure}\par ~~~ \leavevmode \psfig{figure=ds1752f12.eps,width=160mm}\end{figure}$

Figure 12: A pole-on view of the photometric model ( left) and radar profile ( right) of 1627 Ivar. The small ticks correspond to the uncertainties of the fit

**Table 6:** The determined periods, amplitudes, spin vectors and shapes
Asteroid	$P_{{\rm sin}} (h)$	$P_{{\rm sid}} (d)$	A (mag)	${\lambda _{\rm p}}$	${\beta _{\rm p}}$	a/b	b/c	method
683	4.6		0.13	15/195 $\pm$ 25	52 $\pm$ 15	1.15 $\pm$ 0.05	1.05 $\pm$ 0.05	A
		0 $.\!\!^{\rm d}$ 1964156 R						O-C
725	$\geq$ 3		$\geq$ 0.4	-	-	-	-	A
852	4.62		0.32, 0.27	30/210 $\pm$ 20	30 $\pm$ 10	2.3 $\pm$ 0.3	1.2 $\pm$ 0.2	A
1627	4.80		0.77, 0.92	145/325 $\pm$ 8	34 $\pm$ 6	2.0 $\pm$ 0.1	1.09 $\pm$ 0.05	A
		0 $.\!\!^{\rm d}$ 1999154 R		143	-37			O-C
1998 PG	2.6		0.09					Fourier
--	5.3		0.08					Fourier

The Near Earth Object (NEO) 1998 PG was discovered by the LONEOS project in Flagstaff, on August 3, 1998. We observed about 80 days after the discovery, in October, 1998. We found complex, strongly scattering lightcurves (two of them are shown in Figs. 14-15), which did not show any usual regularity. Therefore, we performed a conventional frequency analysis by calculating Discrete Fourier Transform (DFT) of the whole dataset (Fig. 16). Data obtained on October 27 are too noisy, thus we excluded them from the period determination.

The determined periods are 1 $.\!\!^{\rm h}$ 3 and 5 $.\!\!^{\rm h}$ 3, although these values have large uncertainties (about 10-15%). Assuming that the shorter period is due to rotation, we get a rotational period of 2 $.\!\!^{\rm h}$ 6. We note that our period values do not contradict those obtained by P. Pravec and his collaborators, who found $P_{{\rm rot}}$ = 2 $.\!\!^{\rm h}$ 517 and $P_2\approx$ 7 $.\!\!^{\rm h}$ 0 (Pravec 1998, personal communication). The reason for doubly periodic lighcurve can be precession and/or binarity. The observed rate of multiperiodic lightcurves among NEOs is quite high (see, e.g., Pravec 1999), but the underlying physical processes can only be identified with more detailed observations than we have on 1998 PG. Therefore, we conclude that we may have found evidence for precession in 1998 PG, but other explanations cannot be excluded.

3 Discussion