The current state of our knowledge of the physical parameters and the chemical composition of dense cores in molecular clouds is discussed. In particular, I summarize what is known about the rate at which molecules condense out on grain surfaces. I discuss in turn: a) dense cores in nearby dust complexes such as Taurus, b) clumps in regions of massive star formation such as Orion and M17, c) hot dense cores near to newly formed O stars such as the Orion-KL hot core and d) the high density condensations which give rise to interstellar masers. Recent work on each of these categories is reviewed with emphasis on the chemical abundance determinations and estimates of the local density and temperature. Particular attention is given to recent work on OH, methanol, and ammonia masers.

Our understanding of the molecular phase of the interstellar medium is critically dependent on use of various lines from different molecular species to trace this dense material. As our knowledge of molecular clouds becomes more refined, and we pursue in detail issues of molecular cloud structure, stability, and how star formation depends on and affects the molecular gas, it is appropriate to examine the basis by which we determine the morphology of clouds, their density, and other key parameters. This is obviously a major undertaking, well beyond the scope of the short presentation at this conference, so I will concentrate on one very basic, but critical issue, which is that of abundance variations of tracers of density and molecular column density which are widely used to delineate the denser portions of all types of molecular clouds. In this summary, I will first highlight some of the apparent indications of significant variations of abundance within individual clouds, as a way of indicating some potential dangers and the importance of the molecular tracer selected. I will also briefly suggest how such variations may be themselves important diagnostics of cloud structure and evolution.

The fluid dynamics of MHD shock waves in magnetized molecular gas is reviewed. The different types of shock solutions, and the circumstances under which the different types occur, are delineated. Current theoretical work on C∗- and J-type shocks, and on the stability of C-type shocks, is briefly described. Observations of the line emission from MHD shocks in different regions appear to be in conflict with theoretical expectations for single, plane-parallel shocks. Replacement of plane-parallel shocks by bow shocks may help reconcile theory and observation, but it is also possible that the observed shocks may not be “steady”, or that theoretical models have omitted some important physics.

The dynamics of the cold atomic and molecular gas, on which we focus here, is strongly affected by non equilibrium heating and cooling processes. We give two different examples, in which the breaking of the thermal balance is due respectively to variations of the incident ultraviolet radiation flux, and non equilibrium abundances of H 2 molecules in molecular clouds envelopes. Fluctuations of the ultraviolet radiation flux in clumpy molecular cloud envelopes result in the formation or the destruction of dense regions. Large density contrasts, greater than one order of magnitude, are easily achieved in cloud regions of moderate visual extinction. Condensation or expansion develop on quite short time scales, of the order of a few tenth of million year, and induce collective motions which can feed turbulence.

Another example of the importance of out of equilibrium thermochemical processes is furnished by the study of the H — H 2 transition layers in molecular clouds envelopes. They turn out to be unstable against convection-like motions, driven by the energy released by H 2 photodestruction. The gas velocities involved in these motions are, again, typical of the observed turbulent velocity in clouds envelopes.

The broad line wings in molecular cloud spectra are proposed to result from strong magnetic waves on the periphery of dense cores and in the intercore regions where the Alfvén velocity should be larger than average. The observed line profiles are reproduced by a simple but realistic model, and the ratio of the broad to the narrow line components is found to equal approximately three, independent of cloud parameters, as long as the core/intercore contrast in the local average density is sufficiently large. Interactions between the magnetic waves should produce dense clumps in the non-linear splash regions between converging flows.

In a nonstationary medium (behind the shock front, for instance) the development of isobaric and adiabatic modes of the thermal instability are more preferable. Some examples of the fragmentation of the medium on clouds in case of OH masers are given.

A rotating and slowly-contracting gas cloud is followed in the deep gravitational potential of the galactic center. When the gas density increases as high as (102 — 103)H2 cm—3, which is more than twenty times as large as the background matter density, the self-gravity of the cloud becomes dominant to govern its dynamical structure. The cloud elongates and then splits into two separated objects, as observed at the centers of IC342, NGC6946, and Maffei 2 where we have two symmetric peaks on the major axis of the 12CO (J=1-0) cloud.

The properties and structure of six molecular complexes within 500 pc of the Sun are described and compared. They are generally organized into elongated filaments which appear connected to less elongated, more massive clouds. Their prominent star clusters tend to be located in the massive clouds rather than in the filaments. The complexes have similar structure, but big differences in scale, from a few pc to some 30 pc. They show a pattern of regional virial equilibrium, where the massive, centrally located clouds are close to virial equilibrium, while the less massive filaments and other small clouds have too little mass to bind their observed internal motions. Complexes can be ranked according to increasing size, mass, core mass, and the mass and number of the associated stars: they range from Lupus to Taurus to Ophiuchus to Perseus to Orion B to Orion A. The cores in nearby complexes tend to have maps which are elongated, rather than round. The core size, velocity dispersion, and column density of most cores are consistent with virial equilibrium. Cores in Orion tend to exceed cores in Taurus in their line width, size, temperature, mass, and in the mass of the associated star, if any. Stars in Orion tend to be more numerous and more massive than in Taurus, while those in Taurus tend to be more numerous and more massive than in Lupus. The mass of a core tends to increase with the mass of the cloud where it is found, with the mass of the star cluster with which it is associated, and with its proximity to a star cluster. These properties suggest that complexes and their constituent cores and clusters develop together over time, perhaps according to the depth of the gravitational well of the complex.

About a dozen distinct dense cores have been identified in the Rho Ophiuchi molecular cloud. The properties of these cores are summarized and compared to the properties of cores in the Taurus molecular cloud, a less efficient region of star formation, and in DR21(OH), a more massive region of star formation. The data are consistent with a picture in which more massive clouds have a higher surface density of cores, which in turn are more massive. The adjacent cores in L1689N have been studied with very high resolution; one has formed stars and one never has. The structure of these cores shows a tendency for duplicity of structures from the largest scales (1 pc) to the smallest (50 AU).

Recent observations of sub-mm and far-IR atomic fine structure and molecular rotational lines give evidence that due to the clumpiness of the molecular cloud cores the UV radiation from newly formed stars affects a very large fraction of the cloud material. Direct observations of the clumpy structure in M17 SW allow to derive several parameters of the clump distribution, in particular the clump mass spectrum and the volume filling factor. Implications of these results in regard to star formation are shortly discussed.

Model computations of protostellar evolution depend very strongly on the initial conditions: Fragmentation of massive cloud cores or coagulation of substellar condensations, the physical state of gas and dust (e.g. the formation of ice-mantles and grain coagulation), the presence of magnetic fields and its effect on gas and dust, and the formation of accretion disks as a consequence of an initial angular momentum of the protostellar condensation. The MPIfR bolometer group together with the molecular spectroscopists R. Mauersberger and T.L. Wilson have embarked on a program aimed at the exploration of the earliest evolutionary stages of high- and low-mass star formation. Here follows a brief progress report.

Current evidence suggests that the stellar initial mass function has the same basic form everywhere, and that its fundamental features are (1) the existence of a characteristic stellar mass of order one solar mass, and (2) the existence of an apparently universal power-law form for the mass spectrum of the more massive stars. The characteristic stellar mass may be determined in part by the typical mass scale for the fragmentation of star forming clouds, which is predicted to be of the order of one solar mass. The power-law extension of the mass spectrum toward higher masses may result from the continuing accretional growth of some stars to much larger masses; the fact that the most massive stars appear to form preferentially in cluster cores suggests that such continuing accretion may be particularly important at the centers of clusters. Numerical simulations suggest that forming systems of stars may tend to develop a hierarchical structure, possibly self-similar in nature. If most stars form in such hierarchically structured systems, and if the mass of the most massive star that forms in each subcluster increases as a power of the mass of the subcluster, then a mass spectrum of power-law form is predicted. Some possible physical effects that could lead to such a relation are briefly discussed, and some observational tests of the ideas discussed here are proposed.

We present new observational data of 13CO J=1−0 emission in nearby molecular cloud complexes in Taurus, Ophiuchus, and Orion, and compare star formation efficiencies in the three regions. The 13CO data have been obtained with the 4m millimeter telescope at Nagoya equipped with a SIS receiver having receiver temperature of 23 K in double side band. The molecular distribution is often characterized by highly filamentary, continuous distribution, rather than well defined clumps. Active star formation is taking place almost exclusively in the densest regions in the molecular distributions. The filamentary distributions often show velocity gradients perpendicular to their elongation, suggesting that they are spinning around their major axes. Roles of photoionization and the spinning motion on formation of stars and clouds are discussed as a step toward understanding the mechanism regulating star formation efficiencies.

Clusters of young stars have been found near a number of compact HII regions. These clusters do not show a turnover in the K-band luminosity and are probably several million years old. In L 1641 only moderate clustering tendency has been observed and many sources show signs of extremely young age.

We consider the distribution and nature of the gas and star formation in three nearby molecular cloud complexes: Taurus-Auriga, Ophiuchus, and Orion. Large-scale studies show that quite different distributions of dense gas and star formation exist in these regions, but that the global efficiency of star formation, so far, is about 1% in all of these complexes, very similar to the final efficiencies deduced for open clusters and associations. Furthermore, evidence for differences in the initial mass function among these regions is weak; all three regions may be compatible with the initial mass function for open clusters and field stars. Finally, we consider specific examples of star formation in each complex; the results indicate that the ideas of Adams, Lada, and Shu (1987) are reasonably consistent with the data on L1551 IRS 5, in the Taurus cloud, but that modifications are likely to be needed for IRAS 16293-2422 in Ophiuchus and for NGC2071 in Orion.

The mechanism underlying the formation of cores and larger scale structures in molecular clouds must play a fundamental role in the physics of star formation since young stellar objects are usually found within or very near cores (Myers et al 1987, Beichman et al 1986). The Taurus cloud is an ideal object to study in this regard because of its proximity (160 pc), and because only low mass star formation is presently occurring there. Barnard's (1927) beautiful optical photograph of the region reveals that the obscuring gas and dust has filamentary structure that is comparable to the size of the cloud complex (several 10's of pc). This structure is clearly seen in CO maps of the region as well (eg. Duvert et al 1986) where it is apparent that structure on much larger size scales than cores is common. In addition to the filamentary structure one also observes that there are small dark clouds present such as L1489, L1495, etc.

An account is given of the observational and theoretical properties that characterize young stars of intermediate-mass (2 ≤ M∗ / M⊙ ≤ 10), known as Herbig Ae/Be stars. The mass range of these objects is an interesting one, since it involves the occurrence of complex phenomena associated with the transition from fully convective configurations, typical of T Tauri stars, to radiatively stable objects, as in the case of massive stars. An overview of relevant observations testifying the variety of surface phenomena associated with the Herbig Ae/Be stars is presented. Recent developments in the theory of the formation on intermediate-mass protostars will also be discussed.

We present results from a near-infrared array, CO interferometer, and H 2 O maser interferometer survey of the circumstellar environments of 26 young low-luminosity embedded stars located in nearby molecular clouds. About 75% of the sample show evidence for stellar winds/outflows in the near-infrared or CO data indicating that most of these sources are in the early wind clearing phase of their evolution. Close to 15% are multiple on the scale of 20″, suggesting that fragmentation of their surrounding dense cloud cores is important before or during gravitational collapse. Roughly 10% have H 2 O maser emission and the kinematics imply the masers arise in gravitationally unbound gas (i.e., a stellar wind or outflow) rather than in a circumstellar disk.

The CS (J = 2 — 1) line and 98 GHz continuum emission have been observed for 11 protostellar IRAS sources in the Taurus molecular cloud with resolutions of 2.6″−8.8″ (360 AU—1200 AU) using the Nobeyama Millimeter Array (NMA). The CS emission is detected only toward embedded sources, while the continuum emission from dust grains is detected only toward visible T Tauri stars except for one embedded source, L1551-IRS5. This suggests that the dust grains around the embedded sources do not centrally concentrate enough to be detected with our sensitivity (∼4 m Jy r.m.s), while dust grains in disks around the T Tauri stars have enough total mass to be detected with the NMA. The molecular cloud cores around the embedded sources are moderately extended and dense enough to be detected in CS, while gas disks around the T Tauri are not detected because the radius of such gas disks may be smaller than 70 (50 K/Tex) AU. These results imply that the total amount of matter within the NMA beam size must increase when the central objects evolve into T Tauri stars from embedded sources, suggesting that the compact and highly dense disks around T Tauri stars are formed by the dynamical mass accretion during the embedded protostar phase.

We have made aperture synthesis maps toward the Orion-KL nebula in CS(1-0), CS(2-1) and 3mm continuum with the Nobeyama Millimeter Array (NMA). The CS(1-0) maps were obtained across the central 2′ region of the Orion-KL nebula. Our maps show the rotating disk around the KL nebula with the diameter of 0.3 pc and the mass of 25 M⊙. The disk clearly shows the Keplerian rotation, which indicates that the mass of the KL nebula is 100 — 150 M⊙. The shell structure around the molecular outflow is also found in CS(1-0) map. This structure also appears in CS(2-1) map.

From 3mm continuum map, we have found at least six continuum sources. One of them is f-f emission from the ultra-compact HII region around BN. Other sources are the dust clumps whose masses are estimated to be 1 — 4 M⊙. The distribution of the dust clumps shows clear anti-correlation with the 20 μm emissions. This indicates that the 20 μm emissions are obscured by these dust clumps.