We present an analysis of the compact X-ray source population in the Seyfert 2 galaxy NGC 1068 , imaged with a \sim 50 ks Chandra 
observation .
We find a total of 84 compact sources on the S3 chip , of which 66 are located within the 25.0 B-magnitude ( arc sec ) ^ { -2 } 
isophote of the galactic disk of NGC 1068 .
Spectra have been obtained for the 21 sources with at least 50 counts , and modeled with both multi-color disk blackbody and power-law models .
The power-law model provides the better description of the spectrum for 18 of these sources .
For fainter sources , the spectral index has been estimated from the hardness ratio .
Five sources have 0.4 - 8 keV intrinsic luminosities greater than 10 ^ { 39 } erg s ^ { -1 } , assuming that their emission is isotropic and that they are associated with NGC 1068 .
We refer to these sources as Intermediate Luminosity X-ray Objects ( IXOs ) .
If these five sources are X-ray binaries accreting with luminosities that are both sub-Eddington and isotropic , then the implied source masses are \mathrel { \hbox { \raise 2.15 pt \hbox { $ > $ } \hbox to 0.0 pt { \lower 2.15 pt \hbox { $ \sim$% } } } } 7 M _ { \odot } , and so they are inferred to be black holes .
Most of the spectrally modeled sources have spectral shapes similar to Galactic black hole candidates .
However , the brightest compact source in NGC 1068 has a spectrum which is much harder than that found in Galactic black hole candidates and other IXOs .
The brightest source also shows large amplitude variability on both short-term and long-term timescales , with the count rate possibly decreasing by a factor of two in \sim 2 ks during our Chandra observation , and the source flux decreasing by a factor of five between our observation and the grating observations taken just over nine months later .
The ratio of the number of sources with luminosities greater than 2.1 \times 10 ^ { 38 } erg s ^ { -1 } in the 0.4 – 8 keV band to the rate of massive ( > 5 M _ { \odot } ) star formation is the same , to within a factor of two , for NGC 1068 , the Antennae , NGC 5194 ( the main galaxy in M51 ) , and the Circinus galaxy .
This suggests that the rate of production of X-ray binaries per massive star is approximately the same for galaxies with currently active star formation , including “ starbursts ” .