Using infrared data from the Herschel Space Observatory and Karl G. Jansky Very Large Array (VLA) 3 GHz observations in the COSMOS field, we investigate the redshift evolution of the infrared-radio correlation (IRRC) for star-forming galaxies (SFGs) we classify as either spheroid- or disc-dominated based on their morphology. The sample predominantly consists of disc galaxies with stellar mass ${\gtrsim}10^{10}\,M_{\odot}$, and residing on the star-forming main sequence (MS). After the removal of AGN using standard approaches, we observe a significant difference between the redshift-evolution of the median IR/radio ratio $\overline{q}_{\mathrm{TIR}}$ of (i) a sample of ellipticals, plus discs with a substantial bulge component (`spheroid-dominated' SFGs) and, (ii) virtually pure discs and irregular systems (`disc-dominated' SFGs). The spheroid-dominated population follows a declining $\overline{q}_{\mathrm{TIR}}$ vs. $z$ trend similar to that measured in recent evolutionary studies of the IRRC. However, for disc-dominated galaxies, where radio and IR emission should be linked to star formation in the most straightforward way, we measure very little change in $\overline{q}_{\mathrm{TIR}}$. This suggests that low-redshift calibrations of radio emission as an SFR-tracer may remain valid out to at least $z\,{\simeq}\,1\,{-}\,1.5$ for pure star-forming systems. We find that the different redshift-evolution of $q_{\rm TIR}$ for the spheroid- and disc-dominated sample is mainly due to an increasing radio excess for spheroid-dominated galaxies at $z\,{\gtrsim}\,$0.8, hinting at some residual AGN activity in these systems. This finding demonstrates that in the absence of AGN the IRRC is independent of redshift, and that radio observations can therefore be used to estimate SFRs at all redshifts for genuinely star-forming galaxies.

We study the composition of the faint radio population selected from the VLA-COSMOS 3GHz Large Project, a radio continuum survey performed at 10 cm wavelength. The survey covers the full 2 square degree COSMOS field with mean $rms\sim2.3$ $\mu$Jy/beam, cataloging 10,899 source components above $5\times rms$. By combining these radio data with UltraVISTA, optical, near-infrared, and Spitzer/IRAC mid-infrared data, as well as X-ray data from the Chandra Legacy, and Chandra COSMOS surveys, we gain insight into the emission mechanisms within our radio sources out to redshifts of $z\sim5$. From these emission characteristics we classify our souces as star forming galaxies or AGN. Using their multi-wavelength properties we further separate the AGN into sub-samples dominated by radiatively efficient and inefficient AGN, often referred to as high- and low-excitation emission line AGN. We compare our method with other results based on fitting of the sources' spectral energy distributions using both galaxy and AGN spectral models, and those based on the infrared-radio correlation. We study the fractional contributions of these sub-populations down to radio flux levels of $\sim$10 $\mu$Jy. We find that at 3 GHz flux densities above $\sim$400 $\mu$Jy quiescent, red galaxies, consistent with the low-excitation radio AGN class constitute the dominant fraction. Below densities of $\sim$200 $\mu$Jy star-forming galaxies begin to constitute the largest fraction, followed by the low-excitation, and X-ray- and IR-identified high-excitation radio AGN.

Recent studies have found that radio-AGN selected by radio-loudness show little difference in terms of their host galaxy properties when compared to non-AGN galaxies of similar stellar mass and redshift. Using new 1.4~GHz VLBI observations of the COSMOS field we find that approximately 49$\pm8$\% of high-mass (M $>$ 10$^{10.5}$ M$_{\odot}$), high luminosity (L$_{1.4}$ $>$ 10$^{24}$ W~Hz$^{-1}$) radio-AGN possess a VLBI detected counterpart. These objects show no discernible bias towards specific stellar masses, redshifts or host properties other than what is shown by the radio-AGN population in general. Radio-AGN that are detected in VLBI observations are not special, but form a representative sample of the radio-loud AGN population.

(Abridged) Fossil systems are group- or cluster-sized objects whose luminosity is dominated by a very massive central galaxy. In the current cold dark matter scenario, these objects formed hierarchically at an early epoch of the Universe and then slowly evolved until present day. That is the reason why they are called {\it fossils}. We started an extensive observational program to characterize a sample of 34 fossil group candidates spanning a broad range of physical properties. Deep $r-$band images were taken for each candidate and optical spectroscopic observations were obtained for $\sim$ 1200 galaxies. This new dataset was completed with SDSS DR7 archival data to obtain robust cluster membership and global properties of each fossil group candidate. For each system, we recomputed the magnitude gaps between the two brightest galaxies ($\Delta m_{12}$) and the first and fourth ranked galaxies ($\Delta m_{14}$) within 0.5 $R_{{\rm 200}}$. We consider fossil systems those with $\Delta m_{12} \ge 2$ mag or $\Delta m_{14} \ge 2.5$ mag within the errors. We find that 15 candidates turned out to be fossil systems. Their observational properties agree with those of non-fossil systems. Both follow the same correlations, but fossils are always extreme cases. In particular, they host the brightest central galaxies and the fraction of total galaxy light enclosed in the central galaxy is larger in fossil than in non-fossil systems. Finally, we confirm the existence of genuine fossil clusters. Combining our results with others in the literature, we favor the merging scenario in which fossil systems formed due to mergers of $L^\ast$ galaxies. The large magnitude gap is a consequence of the extreme merger ratio within fossil systems and therefore it is an evolutionary effect. Moreover, we suggest that at least one candidate in our sample could represent a transitional fossil stage.