We have used VLBA fringe visibility data obtained at 15 GHz to examine the compact structure in 250 extragalactic radio sources. For 171 sources in our sample, more than half of the total flux density seen by the VLBA remains unresolved on the longest baselines. There are 163 sources in our list with a median correlated flux density at 15 GHz in excess of 0.5 Jy on the longest baselines. For about 60% of the sources, we have at least one observation in which the core component appears unresolved (generally smaller than 0.05 mas) in one direction, usually transverse to the direction into which the jet extends. BL Lacs are on average more compact than quasars, while active galaxies are on average less compact. Also, in an active galaxy the sub-milliarcsecond core component tends to be less dominant. IDV sources typically have a more compact, more core-dominated structure on sub-milliarcsecond scales than non-IDV sources, and sources with a greater amplitude of intra-day variations tend to have a greater unresolved VLBA flux density. The objects known to be GeV gamma-ray loud appear to have a more compact VLBA structure than the other sources in our sample. This suggests that the mechanisms for the production of gamma-ray emission and for the generation of compact radio synchrotron emitting features are related. The brightness temperature estimates and lower limits for the cores in our sample typically range between 10^11 and 10^13 K, but they extend up to 5x10^13 K, apparently in excess of the equipartition brightness temperature, or the inverse Compton limit for stationary synchrotron sources. The largest component speeds are observed in radio sources with high observed brightness temperatures, as would be expected from relativistic beaming (abridged).Comment: 31 pages, 13 figures, 4 tables, accepted for publication in the Astronomical Journal; minor changes to the text are mad

We report the results of a 15 GHz (2 cm) multi-epoch VLBA program, begun in 1994 to study the outflow in radio jets ejected from quasars and active galaxies. The observed flow of 208 distinct features measured in 110 quasars, active galaxies, and BL Lac objects shows highly collimated relativistic motion with apparent transverse velocities typically between zero and about 15c, with a tail extending up to about 34c. Within individual jets, different features appear to move with a similar characteristic velocity which may represent an underlying continuous jet flow, but we also see some stationary and even apparently inward moving features which co-exist with the main features. Comparison of our observations with published data at other wavelengths suggests that there is a systematic decrease in apparent velocity with increasing wavelength, probably because the observations at different wavelengths sample different parts of the jet structure. The observed distribution of linear velocities is not consistent with any simple ballistic model. Either there is a rather broad range of Lorentz factors, a significant difference between the velocity of the bulk relativistic flow and the pattern speed of underlying shocks, or a combination of these options. Assuming a ballistic flow, comparison of observed apparent velocities and Doppler factors computed from the time scale of flux density variations is consistent with a steep power law distribution of intrinsic Lorentz factors, an isotropic distribution of orientations of the parent population, and intrinsic brightness temperatures about an order of magnitude below the canonical inverse Compton limit. It appears that the parent population of radio jets is not dominated by highly relativistic flows (abridged).Comment: 25 pages, 11 figures, 3 tables, accepted for publication in the Astrophysical Journa

We have used the VLBA at 15 GHz to image the structure of 132 strong compact AGN and quasars with a resolution better than one milliarcsecond and a dynamic range typically exceeding 1000 to 1. These observations were made as part of a program to investigate the sub-parsec structure of quasars and AGN and to study the changes in their structure with time. Many of the sources included in our study, particularly those located south of +35 degrees, have not been previously imaged with milliarcsecond resolution. Each of the sources has been observed at multiple epochs. In this paper we show images of each of the 132 sources which we have observed. For each source we present data at the epoch which had the best quality data. The milliarcsecond jets generally appear one-sided but two-sided structure is often found in lower luminosity radio galaxies and in high luminosity quasars with gigahertz peaked spectra. Usually the structure is unresolved along the direction perpendicular to the jet, but a few sources have broad plumes. In some low luminosity radio galaxies, the structure appears more symmetric at 2 cm than at long wavelengths. The apparent long wavelength symmetry in these sources is probably due to absorption by intervening material. A few sources contain only a single component with any secondary feature at least a thousand times weaker. We find no obvious correlation of radio morphology and the detection of gamma-ray emission by EGRET.Comment: 19 pages, 3 tables, 3 figures. Figure 2 (132 contour diagrams) is long and is omitted here. Figure 2 may be viewed at http://www.cv.nrao.edu/2cmsurvey/ In press, Astronomical Journal, April 199

Relations between the observed quantities for a beamed radio jet, apparent transverse speed and apparent luminosity (beta_app,L), and the intrinsic quantities, Lorentz factor and intrinsic luminosity (gamma,L_o), are investigated. The inversion from measured to intrinsic values is not unique, but approximate limits to gamma and L_o can be found using probability arguments. Roughly half the sources in a flux density--limited, beamed sample have a value of gamma close to the measured beta_app. The methods are applied to observations of 119 AGN jets made with the VLBA at 15 GHz during 1994-2002. The results strongly support the common relativistic beam model for an extragalactic radio jet. The (beta_app,L) data are closely bounded by a theoretical envelope, an aspect curve for gamma=32, L_o= 10^25 W/Hz. This gives limits to the maximum values of gamma and L_o in the sample: gamma_max about 32, and L_o,max ~ 10^26 W/Hz. No sources with both high beta_app and low L are observed. This is not the result of selection effects due to the observing limits, which are flux density S>0.5 Jy, and angular velocity mu<4 mas/yr. Many of the fastest quasars have a pattern Lorentz factor gamma_p close to that of the beam, gamma_b, but some of the slow quasars must have gamma_p<