Winzer (1974) found HR 6958 was a photometric variable. He used HR 6985 as the comparison star and found a period of 0.9451 days. Burke & Barr (1981) obtained additional UBV photometry with HD 1702999 and HD 171802 (= HR 6985) as the comparison stars and found a similar period. Spectrograms show it is relatively sharp-lined which is inconsistent with a photometric period of this length.
In a first attempt at using FCAPT data, I could not determine which period was
significant and so began to obtain at least two observations per night. A
total of 379 uvby observations (Table 4) were obtained during the first
five years of observations. A periodogram analysis of the u values
indicated that the most likely period was 18.072 days. Winzer's period
corresponds to the frequency of this period plus 1.0. Examination of
photometry obtained on the same night and the sharp-lined nature of this
star argued for the longer period. As the amplitude of HR 6958 is greatest
in u, constants were added to the U photometry of Winzer
(1974) and of Burke & Barr (1981) so that their mean
values were the same as that for u. When these values were plotted with
the u data, small adjustments in both the period and the epoch of the
maximum value for u are needed. Thus 
As both the maxima and minima are asymmetric there is no advantage of
selecting one over the other for determining the zero phase.
Figure 3 (click here) shows the uvby values plotted according to the adopted ephemeris. In the panel for u, the transformed U values of Winzer (1974) and Burke & Barr (1981) are plotted as close diamonds and open circles, respectively. The FCAPT values are always plotted as plus signs. Maxima in different colors appear to occur at slightly different phases with that for v occurring earlier than that for u and those for b and y occurring slightly later than that for u. The amplitudes of variation are 0.03 mag, 0.02 mag, 0.02 mag, and 0.015 mag, respectively for u, v, b, and y. The light curve for u is quite asymmetric with the rising branch of the light curve taking about 40% of the period. That for v is somewhat similar to that for u. But those of b and y are nearly symmetric.