Abstract
In 1958 Jan Oort remarked that the lack of old clusters in the solar neighborhood (SN) implies that clusters are destroyed on a timescale of less than a Gyr. This is much shorter than the predicted dissolution time of clusters due to stellar evolution and two-body relaxation in the tidal field of the Galaxy. So, other (external) effects must play a dominant role in the destruction of star clusters in the solar neighborhood. We recalculated the survival time of initially bound star clusters in the solar neighborhood taking into account: (1) stellar evolution, (2) tidal stripping, (3) perturbations by spiral arms and (4) encounters with giant molecular clouds (GMCs). We find that encounters with GMCs are the most damaging to clusters. The resulting predicted dissolution time of these combined effects, t_dis=1.7 (Mi/10^4 M_sun)^0.67 Gyr for clusters in the mass range of 10^2 < M < 10^5 M_sun, is very similar to the disruption time of t_dis=1.3+/-0.5 (M/10^4 M_sun)^0.62 Gyr that was derived empirically from a mass limited sample of clusters in the solar neighborhood within 600 pc. The predicted shape of the age distribution of clusters agrees very well with the observed one. The comparison between observations and theory implies a surface star formation rate (SFR) near the sun of 3.5x10^-10 M_sun yr^-1 pc^-2 for stars in bound clusters with an initial mass in the range of 10^2 to 3x10^4 M_sun. This can be compared to a total SFR of 7-10x10^-10 M_sun yr^-1 pc^-2 derived from embedded clusters or 3-7x10^-9 M_sun yr^-1 pc^-2 derived from field stars. This implies an infant mortality rate of clusters in the solar neighborhood between 50% and 95%, in agreement with the results of a study of embedded clusters.