Photometric observations of MT Ser, the central star of the planetary nebul
a Abell 41 are presented. The periodic modulations detected by Grauer & Bon
d (1983) are confirmed, thus firmly establishing the binary nature of MT Se
r. The significantly enlarged time base permits us to derive more accurate
ephemeris. The orbital period is either P-1 = 0.113226533 days or twice tha
t value, P-2 = 0.226453066 days. We analyze the light curve (after a carefu
l subtraction of the nebular contribution) with the Wilson-Devinney light c
urve synthesis routine. Since it is not a priori clear which is the true or
bital period of MT Ser, two radically different models, one based on P-1 th
e other on P-2 are considered: (1) A low temperature component orbiting aro
und a hot sub-dwarf. The variability is then due to a reflection effect tog
ether with ellipsoidal variations of one or both components. (2) Two hot su
b-dwarfs of similar temperature and luminosity, partially eclipsing each ot
her and exhibiting ellipsoidal variations. In both models, the primary as w
ell as the secondary component are required to almost fill their respective
Roche lobes. A contact configuration is possible. Pros and cons can be fou
nd for either of the two models. A final decision between them has to await
the observations of a radial velocity curve. The orbital period is current
ly decreasing at a rate of (P) over dot/P = -1.15 x 10(-9) d(-1). Interpret
ing this as due to mass loss via a stellar winds permits us to estimate mas
s loss rates depending on the different model assumptions.