Methods used in pattern recognition and cluster analysis are applied to investigate the spatial distribution of the star forming regions. The fractal dimension of these structures is deduced. The new 21 cm, radio continuum (1.4 GHz) and IRAS surveys reveal scale structures of 700 pc to 1500 pc being identical with the optically identified star forming sites. The morphological structures delineated by young stars reflect physical parameters which determine the star formation in this galaxy. The formation of spiral arm filaments is understandable by stochastic selfpropagating star formation processes.

Results from a continuing study (Figure 1a) of the CO (J = 1-0) distribution in M51 are presented. The angular resolution (beam size 33″ = 1.5 kpc) and pointing stability (better than 4″) of the Onsala 20-m telescope have extracted a number of new features of the molecular cloud distribution and kinematics of M51 [1,2]:

A radio continuum survey of the galactic plane has been made with the 45-m telescope of the Nobeyama Radio Observatory at 10.55 GHz, which is the highest frequency among such surveys. The sensitivity of the telescope was Tb/S = 0.47 K/Jy and the HPBW was 2!6, which was a great advantage because of the same beam size of the Bonn 5-GHz survey (Altenhoff et al. 1978). The receiver was a cooled parametric amplifier. The instantaneous bandwidth was 500 MHz, and the system noise temperature was about 100 K. The calibration source was NGC 7027, which was assumed to be 6.6 Jy. One circular polarization component was observed. The observational parameters are summarized in Table I.

We performed an N-body simulation of a system of gravitating particles modeling the system of stable Giant Molecular Clouds (GMCs) in the background density wave potential. The system of gravitating GMCs forms shocks in the spiral potential well, and shows a prominent spiral structure.

The formation process of giant molecular clouds (GMCs) is investigated from the standpoint of the coagulation theory of molecular clouds. Small clouds collide with each other and grow to become massive ones. Ultimately they form GMCs with a finite lifetime. The occurrence of star formation in a GMC destroys it and consequently small clouds are formed again. We study the time evolution of the clouds which move through a spiral gravitational potential by an N-body simulation. Then the ensemble of clouds responds to the spiral potential and forms a spiral structure similar to that produced by hydrodynamical galactic shock. It is shown that GMCs are formed in the spiral arm region by collisions between clouds. The distribution of GMCs indicates their short lifetime, of the order of a few times 107 years.

There are very good reasons to believe that every stellar clustering in the Milky Way is a star-forming region or at least that every association is such a region and the minor clusterings adjacent sub-regions.

Some star-forming clusterings are still active but it seems as if a majority should be burned-out. The problem is to estimate the total number of clusterings in the present Milky Way, de-activated and disintegrated as well as active.

In addition to the well-known radial abundance gradient in the disc of our Galaxy, there are indications of small-scale variations (Talent and Dufour 1979, Giridhar 1983). We have compiled spectroscopic abundances for long period Cepheids and to eliminate the effect of differences in ages, derived their places of formation by calculating their galactic orbits backwards in time. The stellar orbits were calculated considering the axisymmetric gravitational potential as well as the spiral potential due to the density wave. The parameters related to spiral waves are taken from Yuan (1969). The equations of motion in the (ζ, η, ξ) system were numerically integrated using the Runge-Kutta method and birthsites were transformed to (ζ, η) in the reference frame of the spiral pattern. The details of the computations are described by Giridhar (1985). In addition to the accepted value of 13.5 km s−1 kpc−1 for pattern velocity, birthsites were also calculated for ωp = 11.5 and 15.5 km s−1 kpc−1. Figure 1 shows the birthsites in the galactic plane for the three values of the pattern velocity. The Cepheids in the figure can be identified through the serial number in Table I which contains our compilation of Fe/H and derived birthsites. It is obvious from the figure that the systematic shifts in birthsites for different pattern velocities are not large enough to affect the assignment of birthsites to different arms.

In attempting to reproduce the initial stellar mass function, we solved analytically the coagulation equation with an explicit time dependence in the coagulation rate in order to simulate the gravitational collapse of the fragments upon themselves as they move within the progenitor cloud. Two separate cases have been studied, with and without a mass dependence in the coagulation rate. The solution show that (1) inclusion of self-gravitation is very important and can change the results to the point of preventing coalescence to work altogether, depending on the values of the two free parameters, (2) the precise form of the mass dependence of the coagulation rate is not of prime importance in most situations of astrophysical interest, and (3) coagulation alone is not sufficient to yield a realistic mass spectrum and fragmentation must also be taken into account. Coagulation is more efficient for massive fragments and fragmentation for the smaller ones. These results are applied to different regions: star clusters, associations, and starburst regions.

To model the nearly constant star formation rate, SFR, observed in most late-type galaxies (Gallagher et al. 1984), as well as the starbursting behaviour of some interacting galaxies (Joseph et al. 1984) we have developed a phenomenological model of the regulatory coupling between the density of the star forming part of the interstellar medium, i.e. molecular gas, and the rate of formation of massive stars.

We present the results of an extensive survey of 13CO and CS emission in the inner 1 kpc region of our Galaxy. The properties of the clouds are very different from molecular clouds observed in the galactic disc: 1) The mean gas density is more than one order of magnitude higher (<n> = 104 cm−3). 2) The internal velocity dispersion is greater by a factor of 5 to 10 (v = 20 to 50 km s−1). 3) Many of the clouds follow highly eccentric orbits about the center. 4) This region contains roughly 10% of the Galaxy's molecular mass, but much less than 10% of the newborn massive stars. Star formation may be inhibited by the large turbulent energy content of the clouds. Some of these properties may result from the large galactic tides in the inner Galaxy. In order for a cloud to remain gravitationally bound despite the shear induced by galactic rotation, the mean gas density must be greater than <n> = 104 cm−3. The velocity dispersion of the gas must be large to satisfy the virial theorem. The forbidden velocities and high latitude extent of some of the emission can be explained by the presence of a stellar bar and does not require a central explosion. Some of the thermal infrared and continuum features of the galactic center can be understood as the direct consequence of cloud-cloud interactions.

While much work has been devoted to the history of star formation in the Magellanic Clouds at large spatial and time scales, little has been done on the details of sequential formation at the scale of a few 106 or 107 years and of less than 1000 pc.

We present the results of a search for compact infrared objects in the direction of type I OH/H2O maser sources discovered towards H II regions in the Large Magellanic Cloud.

The problem of stellar population present in a young stellar system is studied from the point of view of the evolutionary synthesis. The dependence of the observable properties on: a) stellar mass loss, b) initial mass function, c) star formation rate, and d) age of stellar population is explored. Properties, such as colors and number of lyman continuum photons, that vary rapidly in time are derived and compared with existing data. The range from the UV to the IR is covered in our predictions. We find good agreement between the observations and our predictions for reasonable values of our free parameters.

The distance to a star forming region can be determined by measuring the proper motions within H2O maser clusters. If the motions of the maser spots are random, the distance can be determined by applying the technique known as statistical parallax. Alternatively, if organized motions are evident in the proper motions, one can model the source to estimate its the distance. Both methods rely on a comparison of the radial component of the motion (in km/s) and the proper motion on the plane of the sky (in milli-arcseconds/year).

An infrared complex has been found in the radio arc region near the Galactic center. The complex consists of three sources that are close (< 10″) to each other, and are almost identical in every point of their characteristics; having the same energy spectrum and the same polarization. The observed polarizations are large; 5% at the K-band, and are parallel to the galactic plane. Both behaviors are compatible to those of the galactic center sources, suggesting that the sources are located near the galactic center. The energy spectra are very similar to each other, with large infrared excesses, peaking near the M-band. The luminosity of each source is estimated to be as high as 3-5x105 L⊙, after correcting for interstellar extinction assuming that they are located near the Galactic center; their luminosity is comparable to those of supergiant stars. By CVF spectrophotometry no CO-band absorption nor Brγ emission has been detected, thus no evidence for either M-supergiant nor OB supergiant has been obtained. On the other hand, the very close linear distances, 0.5 pc among each other, suggests their physical relationship, i.e., they should be very young objects, otherwise they would have been dispersed far apart.

Observations of the molecular cloud distributions in spiral galaxies are reviewed. For the luminous, relatively face-on Sc galaxies, the azimuthally averaged CO distributions are centrally peaked, with H2 surface densities which decrease as a function of radius. For the Sb and Sa galaxies, the CO distributions exhibit central CO holes up to 5 kpc across in a significant fraction of the galaxies studied. In galaxies with this CO morphology, the central hole is coincident with the nuclear bulge of the galaxy. Additionally, the radial distributions of CO and 13CO emission are similar in 10 Sb and Sc galaxies.

A minority of spiral galaxies is currently undergoing bursts of intense star formation. This paper reviews studies that have been made of these galaxies at visible, radio, and infrared wavelengths, with an emphasis on results that have emerged since the IRAS survey.

Clumpy irregular galaxies have been identified as giant irregulars with a clumpy structure. Investigations in the visible, UV, radio centimetric, X-ray and far infrared demonstrated that the clumps are hyperactive bursts of star formation, each equivalent to a hundred giant HII regions like 30 Doradus or NGC 604. In spite of their strong activity, their linear size is smaller than half a kpc, leading to an “olive jar” model with very peculiar physical conditions. These could be related to the non-detection of the 2.6-mm CO line and to our discovery in a clump of the first known case of a strong, compact, variable radio source which is not a galactic nucleus.

Normal, non-interacting irregular galaxies can be quite successful at forming stars. Therefore, spiral density waves are not necessary to a vigorous production of stars. Nevertheless, there is a large range in star-formation rates among irregular galaxies. Irregulars with common characteristics can have different overall levels of star-formation activity, so that the level of activity does not seem to be simple related to observable global properties of galaxian systems. The constant star formation rates of most normal irregulars also imply the existence of regulatory processes.

Examination of IRAS results for a sample of nearby galaxies with RC2 types between SO/a and Scd (inclusive) suggests that the only galaxies in which the IRAS colours are dominated by emission from warm dust in star formation regions are barred (Hawarden et al. 1985). Figure 1 is a 2-colour diagram of all Shapley-Ames spiral galaxies of the above types detected by IRAS in all four wavebands. It shows the ratio of fluxes in the IRAS 25 and 12 micron bands as a function of the ratio of fluxes in the 100 and 25 micron bands. On this diagram the SA systems, with the exception of a few Seyferts, have small 25/12 ratios and large 100/25 ratios. In sharp contrast, a large fraction of the SB galaxies and many SAB galaxies (not shown) have greatly enhanced 25 micron fluxes indicative of large amounts of dust at temperatures such as are commonly observed in regions of current star formation.