Understanding the origin of galaxy morphologies is one of the leading challenges in galaxy formation studies. Observations reveal that the spheroid-disk dichotomy is already established in galaxy morphologies of the early Universe, with the existence of two markedly different populations: very extended and rotationally supported disks versus quiescent and very compact massive spheroids. We use state-of-the-art cosmological N-body/gasdynamical simulations drawn from the OWLS (OverWhelmingly Large Simulations) project in order to investigate the mass and spin assembling of galaxies at z=2. I will focus on the effects induced by galactic outflows on the final sizes and angular momentum content of simulated galaxies, and contrast them with the latest results from SINFONI/VLT, OSIRIS/KECK and HST/NIC2 on z~2 objects.

The Lyman-alpha (Lya) emission line has proven the most powerful tool by which to detect high-redshift galaxies, thanks to its high equivalent width and its convenient rest-frame wavelength. It is now widely used for galaxy evolution and star formation diagnostics. However, this line undergoes a complex radiation transport leading to doubtful interpretation of Lya observations. For the plethora of Lya-based cosmological promises to be realized, there are a number of apparent inconsistencies that first need to be understood

After reionization, most of the intergalactic hydrogen-ionizing photons were absorbed in dense regions called Lyman-limit systems. These systems also played an important role in the reionization process. Recently, observational constraints on these systems have improved significantly at z~4, but we still do not have a solid theoretical picture for what they were. Cosmological simulations do excellently at capturing the statistics of lower density intergalactic absorption systems (and even are fairly successful for the much higher density damped Lyman-alpha systems), but previous studies of Lyman-limits have found a factor of 10 under-abundance in simulations compared with observations. In contrast to these prior studies, we find that modern high-resolution SPH simulations with an improved treatment of the intergalactic radiation do remarkably well in reproducing the observed column-density distribution of z>3 Lyman-limits. I will discuss the properties of these systems, the relevant physics that shapes them, and how robust our predictions are to different astrophysical uncertainties. Our results have interesting implications for the feedback of galaxies on their surrounding environment, how many ionizing photons per baryon were needed to reionize the Universe, and the structure of reionization. There are also some nice, intuitive analytic models used to describe the abundance and properties of dense absorption systems that have not been rigorously tested. I will comment on the successes and failures of these models.

Observations reveal, remarkably, that the vast diversity of star formation activities in the Universe can be reproduced with a simple power-law scaling of the star formation rate density on the cold gas density-- the Schmidt law. The combination of a Schmidt law with a low density threshold provides a powerful recipe for modelling and simulating star formation in galaxies, and hints at a simple underlying physical regulating mechanism for the formation of molecular clouds and stars in galaxies. A wide range of observational and theoretical studies are now testing the applicability limits of the Schmidt law and exploring its physical basis. For the first time the elements of a truly physical model for star formation on galactic scales are beginning to emerge. This lecture will review the basics of the star formation vs gas density relation, and highlight the wealth of recent work on the subject.

Over the past few years, early-type galaxies have shed their “red and dead” moniker, thanks to the discovery that many host low-level residual star formation. As part of the ATLAS3D project we are conducting a complete, volume limited survey of the molecular gas in 260 local early-type galaxies with the IRAM-30m telescope and the CARMA interferometer, in an attempt to understand the fuel powering this star formation. We find that around 23% of early-type galaxies in the local volume host gas reservoirs, with molecular central discs, polar structures and rings being common. This detection rate is independent of galaxy luminosity and environment, but does depend on the galaxy kinematics. The origin of the molecular gas seems to depend strongly on environment, with misaligned gas being common in the field but completely absent in Virgo. I will discuss the kinematics and star-formation properties of the CO detected galaxies, and the physical conditions within the molecular gas revealed by dense/optically thin tracers such as 13CO, HCN and HCO+. I will touch on the implications of these results in the context of understanding star formation processes within the red sequence. I will also present the first molecular gas Tully-Fisher relation for early-type galaxies, and show that molecules may be the kinematic tracer of choice for probing the M/L evolution of galaxies over cosmic-time.

In rare cases, gravitational lensing makes galaxies appear 10-30x brighter than they really are. This allows us to make measurements that are normally beyond the normal reach of even the largest telescopes. These observations show galaxies forming stars, consuming gas falling in from the "cosmic web", and synthesizing elements which these galaxies then cast into the voids -- they are Rosetta stones for understanding the evolution of galaxies through cosmic time, including how our own Galaxy came to be. I will summarize my program to obtain multiwavelength, diagnostic spectra for samples of lensed galaxies at redshifts of 1-3. This critical epoch, 2-6 billion years after the Big Bang, is when most of the stars in the Universe formed.

Although observations of ionized and neutral gas outflows in radio-galaxies (RG) suggest that AGN feedback has a galaxy-scale impact on the host ISM, it is still unclear how AGN feedback affects the molecular gas. Studying the physical conditions of the molecular gas in powerful RG is therefore critical if we want to understand how radio sources may regulate the star formation in their host galaxies. Recently, Spitzer IRS observations have revealed that 30% of a sample of 55 nearby 3C RG have luminous mid-IR line emission from warm (100-1000 K) molecular hydrogen (H2), with weak tracers of star formation (e.g. PAHs). Following these results, we obtained deep IRS high-resolution spectroscopy of 8 nearby RG that show fast HI outflows. Strikingly, all of these HI-outflow RG are H2-luminous. This strongly suggests that the radio jet, which drives the HI outflow, is also responsible for the shock-excitation of the warm H2 gas. I will review the peculiar properties of these H2-luminous galaxies, and propose an interpretation based on our recent analysis of the warm and cold multiphase gas content of the 'prototypical' H2 luminous RG, 3C 326 N. A fraction of the mechanical energy of the radio jet is being dissipated within the molecular gas, which powers the observed H2 emission. This is an extension of the classical 'cocoon' model of jet-ISM interaction, explicitly taking into account the multiphase character of the ISM, with an emphasis on the molecular gas. Some of these H2-luminous RG have low star-formation efficiencies, which suggests that these processes may indeed have a significant impact on star formation. Dissipation times suggest that this effect may be long-lasting, and perhaps even exceed the jet lifetime. I will discuss the possible implications of these processes for the evolution of RG in particular, and massive galaxies in general.