We have modeled the emission from dust in pre-protostellar cores, including a self-consistent calculation of the temperature distribution for each input density distribution. Model density distributions include Bonnor-Ebert spheres and power laws. The Bonnor-Ebert spheres fit the data well for all three cores we have modeled. The dust temperatures decline to very low values (T d ∼ 7 K) in the centers of these cores, strongly affecting the dust emission. Compared to earlier models that assume constant dust temperatures, our models indicate higher central densities and smaller regions of relatively constant density. Indeed, for L1544, a power-law density distribution, similar to that of a singular, isothermal sphere, cannot be ruled out. For the three sources modeled herein, there seems to be a sequence of increasing central condensation, from L1512 to L1689B to L1544. The two denser cores, L1689B and L1544, have spectroscopic evidence for contraction, suggesting an evolutionary sequence for preprotostellar cores.

Observations have revealed prodigious amounts of star formation in starburst galaxies as traced by dust and molecular emission, even at large redshifts. Recent work shows that for both nearby spiral galaxies and distant starbursts, the global star formation rate, as indicated by the infrared luminosity, has a tight and almost linear correlation with the amount of dense gas as traced by the luminosity of HCN. Our surveys of Galactic dense cores in HCN 1−0 emission show that this correlation continues to a much smaller scale, with nearly the same ratio of infrared luminosity to HCN luminosity found over 7-8 orders of magnitude in L IR , with a lower cutoff around 10 4.5 L ⊙ of infrared luminosity. The linear correlation suggests that we may understand distant star formation in terms of the known properties of local star-forming regions. Both the correlation and the luminosity cutoff can be explained if the basic unit of star formation in galaxies is a dense core, similar to those studied in our Galaxy.

We have mapped over 50 massive, dense clumps with four dense gas tracers: HCN J = 1 − 0 and 3 − 2; and CS J = 2 − 1 and 7 − 6 transitions. Spectral lines of optically thin H 13 CN 3-2 and C 34 S 5-4 were also obtained towards the map centers. These maps usually demonstrate single well-peaked distributions at our resolution, even with higher J transitions. The size, virial mass, surface density, and mean volume density within a well-defined angular size (FWHM) were calculated from the contour maps for each transition. We found that transitions with higher effective density usually trace the more compact, inner part of the clumps but have larger linewidths, leading to an inverse linewidth-size relation using different tracers. The mean surface densities are 0.29, 0.33, 0.78, 1.09 g cm −2 within FWHM contours of CS 2-1, HCN 1-0, HCN 3-2 and CS 7-6, respectively. We find no correlation of L IR with surface density and a possible inverse correlation with mean volume density, contrary to some theoretical expectations. Molecular line luminosities L ′ mol were derived for each transition. We see no evidence in the data for the relation between L ′ mol and mean density posited by modelers. The correlation between L ′ mol and the virial mass is roughly linear for each dense gas tracer. No obvious correlation was found between the line luminosity ratio and infrared luminosity, bolometric temperature, or the L IR /M V ir ratio. A nearly -2linear correlation was found between the infrared luminosity and the line luminosity of all dense gas tracers for these massive, dense clumps, with a lower cutoff in luminosity at L IR = 10 4.5 L ⊙ . The L IR -L ′ HCN 1−0 correlation agrees well with the one found in galaxies. These correlations indicate a constant star formation rate per unit mass from the scale of dense clumps to that of distant galaxies when the mass is measured for dense gas. These results support the suggestion that starburst galaxies may be understood as having a large fraction of gas in dense clumps.

Fifty-one dense cores associated with water masers were mapped at 350 lm. These cores are very luminous, 10 3 < L bol =L < 10 6 , indicative of the formation of massive stars. Dust continuum contour maps, radial intensity profiles, and photometry are presented for these sources. The submillimeter dust emission peak is, on average, nearly coincident with the water maser position. The spectral energy distributions and normalized radial profiles of dust continuum emission were modeled for 31 sources using a one-dimensional dust radiative transfer code, assuming a power-law density distribution in the envelope, n ¼ n f ðr=r f Þ Àp . The bestfit density power-law exponent, p, ranged from 0.75 to 2.5 with hpi ¼ 1:8 AE 0:4, similar to the mean value found recently by Beuther and coworkers in a large sample of massive star-forming regions. The mean value of p is also comparable to that found in regions forming only low-mass stars, but hn f i is over 2 orders of magnitude greater for the massive cores. The mean p is incompatible with a logatropic sphere (p ¼ 1), but other star formation models cannot be ruled out. Different mass estimates are compared and mean masses of gas and dust are reported within a half-power radius determined from the dust emission, hlog Mð< r dec Þi ¼ 2:0 AE 0:6, and within a radius where the total density exceeds 10 4 cm À3 , hlog Mð< r n Þi ¼ 2:5 AE 0:6. Evolutionary indicators commonly used for low-mass star formation, such as T bol and L bol /L smm , may have some utility for regions forming massive stars. Additionally, for comparison with extragalactic star formation studies, the luminosity-to-dust mass ratio is calculated for these sources, hL bol =M D i ¼ 1:4 Â 10 4 L /M , with a method most parallel to that used in studies of distant galaxies. This ratio is similar to that seen in high-redshift starburst galaxies.

We have obtained 850 and 450 km continuum maps of 21 low-mass cores with SEDs ranging from pre-protostellar to Class I (18 K), using SCUBA at the JCMT. In this paper we present K \ T bol \ 370 the maps, radial intensity proÐles, and photometry. Pre-protostellar cores do not have power-law intensity proÐles, whereas the intensity proÐles of Class 0 and Class I sources can be Ðtted with power laws over a large range of radii. A substantial number of sources have companion sources within a few arcminutes (two out of Ðve pre-protostellar cores, nine out of 16 Class 0/I sources). The mean separation between sources is 10,800 AU. The median separation is 18,000 AU including sources without companions as a lower limit. The mean value of the spectral index between 450 and 850 km is 2.8^0.4, with pre-protostellar cores having slightly lower spectral indices (2.5^0.4). The mean mass of the sample, based on the dust emission in a 120A aperture, is 1.1^0.9For the sources Ðtted by power-law M _ . intensity distributions the mean value of m is 1.52^0.45 for Class 0 and I sources, at 850 km and 1.44^0.25 at 450 km. Based on a simple analysis, assuming the emission is in the Rayleigh-Jeans limit and that these values of m translate into power-law density distribu-T d (r) P r~0.4, tions (n P r~p) with p D 2.1. However, we show that this result may be changed by more careful consideration of e †ects such as beam size and shape, Ðnite outer radii, more realistic and failure of the T d (r), Rayleigh-Jeans approximation.

We present high spatial resolution observations of the multiple protostellar system IRAS 16293−2422 using the Submillimeter Array (SMA) at 300 GHz, and the Very Large Array (VLA) at frequencies from 1.5 to 43 GHz. This source was already known to be a binary system with its main components, A and B, separated by ∼ 5 ′′ . The new SMA data now separate source A into two submillimeter continuum components, which we denote Aa and Ab. The strongest of these, Aa, peaks between the centimeter radio sources A1 and A2, but the resolution of the current submillimeter data is insufficient to distinguish whether this is a separate source or the centroid of submillimeter dust emission associated with A1 and A2. Archival VLA data spanning 18 years show proper motion of sources A and B of 17 mas yr −1 , associated with the motion of the ρ Ophiuchi cloud. We also find, however, significant relative motion between the centimeter sources A1 and A2 which excludes the possibility that these

No abstract

We have mapped 63 regions forming high-mass stars in CS J ¼ 5 ! 4 using the CSO. The CS peak position was observed in C 34 S J ¼ 5 ! 4 toward 57 cores and in 13 CS J ¼ 5 ! 4 toward the nine brightest cores. The sample is a subset of a sample originally selected toward water masers; the selection on maser sources should favor sources in an early stage of evolution. The cores are located in the first and second Galactic quadrants with an average distance of 5:3 AE 3:7 kpc and were well detected with a median peak signalto-noise ratio in the integrated intensity of 40. The integrated intensity of CS J ¼ 5 ! 4 correlates very well with the dust continuum emission at 350 lm. For 57 sufficiently isolated cores, a well-defined angular size (FWHM) was determined. The core radius (R CS ), aspect ratio [ða=bÞ obs ], virial mass (M vir ), surface density (AE), and the luminosity in the CS J ¼ 5 ! 4 line (LðCS54Þ) are calculated. The distributions of size, virial mass, surface density, and luminosity are all peaked with a few cores skewed toward much larger values than the mean. The median values, l 1/2 , are as follows: l 1/2 ðR CS Þ ¼ 0:32 pc, l 1/2 ðða=bÞ obs Þ ¼ 1:20, l 1/2 ðM vir Þ ¼ 920 M , l 1/2 ðAEÞ ¼ 0:60 g cm À2 , l 1/2 ðLðCS54ÞÞ ¼ 1:9 Â 10 À2 L , and l 1/2 ðL bol =M vir Þ ¼ 165 ðL=MÞ . We find a weak correlation between C 34 S line width and size, consistent with Dv $ R 0:3 . The line widths are much higher than would be predicted by the usual relations between line width and size determined from regions of lower mass. These regions are very turbulent. The derived virial mass agrees within a factor of 2-3 with mass estimates from dust emission at 350 lm after corrections for the density structure are accounted for. The resulting cumulative mass spectrum of cores above 1000 M can be approximated by a power law with a slope of about À0.9, steeper than that of clouds measured with tracers of lower density gas and close to that for the total masses of stars in OB associations. The median turbulent pressures are comparable to those in UCH ii regions, and the pressures at small radii are similar to those in hypercompact H ii regions (P=k $ 10 10 K cm À3 ). The filling factors for dense gas are substantial, and the median abundance of CS is about 10 À9 . The ratio of bolometric luminosity to virial mass is much higher than the value found for molecular clouds as a whole, and the correlation of luminosity with mass is tighter.