We present an unbiased census of starless cores in Perseus , Serpens , and Ophiuchus , assembled by comparing large-scale Bolocam 1.1 mm continuum emission maps with Spitzer c2d surveys .
We use the c2d catalogs to separate 108 starless from 92 protostellar cores in the 1.1 mm core samples from Enoch et al .
( 20 ) , Young et al .
( 85 ) , and Enoch et al .
( 21 ) .
A comparison of these populations reveals the initial conditions of the starless cores .
Starless cores in Perseus have similar masses but larger sizes and lower densities on average than protostellar cores , with sizes that suggest density profiles substantially flatter than \rho \propto r ^ { -2 } .
By contrast , starless cores in Serpens are compact and have lower masses than protostellar cores ; future star formation will likely result in lower mass objects than the currently forming protostars .
Comparison to dynamical masses estimated from the NH _ { 3 } survey of Perseus cores by Rosolowsky et al .
( 67 ) suggests that most of the starless cores are likely to be gravitationally bound , and thus prestellar .
The combined prestellar core mass distribution includes 108 cores and has a slope of \alpha = -2.3 \pm 0.4 for M > 0.8 ~ { } \mbox { M$ { } _ { \odot } $ } .
This slope is consistent with recent measurements of the stellar initial mass function , providing further evidence that stellar masses are directly linked to the core formation process .
We place a lower limit on the core-to-star efficiency of 25 % .
There are approximately equal numbers of prestellar and protostellar cores in each cloud , thus the dense prestellar core lifetime must be similar to the lifetime of embedded protostars , or 4.5 \times 10 ^ { 5 } years , with a total uncertainty of a factor of two .
Such a short lifetime suggests a dynamic , rather than quasi-static , core evolution scenario , at least at the relatively high mean densities ( n > 2 \times 10 ^ { 4 } cm ^ { -3 } ) to which we are sensitive .