We have completed an intensive monitoring program of two fields on either side of the center of M31 , and report here on the results concerning microlensing of stars in M31 .
These results stem from a three-year study ( the VATT/Columbia survey ) of microlensing and variability of M31 stars , emphasizing microlensing events of 3 day to 2 month timescales and likely due to masses in M31 .
These observations were conducted intensively from 1997-1999 , with baselines 1995-present , at the Vatican Advanced Technology Telescope and the 1.3-meter telescope at MDM Observatory , with additional data from the Isaac Newton Telescope , including about 200 epochs total .
The two fields monitored cover 560 square arcminutes total , positioned along the minor axis on either side of M31 .
Candidate microlensing events are subject to a number of tests discussed here with the purpose of distinguishing microlensing from variable star activity .
A total of four probable microlensing events , when compared to carefully computed event rate and efficiency models , indicate a marginally significant microlensing activity above that expected for the stars alone in M31 ( and the Galaxy ) acting as lenses .
A maximum likelihood analysis of the distribution of events in timescale and across the face of M31 indicate a microlensing dark matter halo fraction consistent with that seen in our Galaxy towards the Large Magellanic Cloud ( Alcock et al .
2000a ) .
Specifically , for a nearly singular isothermal sphere model , we find a microlensing halo mass fraction f _ { b } = 0.29 ^ { +0.30 } _ { -0.13 } of the total dark matter , and a poorly constrained lensing component mass ( 0.02 to 1.5 M _ { \odot } , 1 \sigma limits ) .
This study serves as the prototype for a larger study approaching completion ; between the two there is significant evidence for an asymmetry in the distribution of microlensing events across the face of M31 , and therefore a large population of halo microlensing dark matter objects .