2000ApJ...541..587D

This paper uses the caustic-crossing events in the microlens data sets to explore the nature and location of the lenses. We conclude that (1) the large majority of lenses, whether they are luminous or dark, are likely to be binaries; (2) blending is an important feature of all the data sets; and (3) most of the LMC/SMC events to date are due to lenses in the Magellanic Clouds, if, as is suspected, the caustic-crossing events along the directions to the Magellanic Clouds are due to lenses located in the Clouds. Although these conclusions are based on a very small number of events, they are subject to a variety of detailed tests. If they are correct, a large fraction of lenses along the direction to the LMC may be ordinary stellar binary systems, just as are the majority of the lenses along the direction to the Bulge. Thus, a better understanding of the larger-than-anticipated value derived for the Bulge optical depth may allow us to better interpret the large value derived for the optical depth to the LMC. Indeed, binarity and blending in the data sets may illuminate connections among several other puzzles: the dearth of binary-source light curves, the dearth of non-caustic-crossing, perturbed, binary-lens events, and the dearth of obviously blended point-lens events.