1: \begin{abstract}

2: 

3: We present two new Lagrangian methods for hydrodynamics, in a systematic comparison with moving-mesh, SPH, and stationary (non-moving) grid methods. The new methods are designed to simultaneously capture advantages of both smoothed-particle hydrodynamics (SPH) and grid-based/adaptive mesh refinement (AMR) schemes. They are based on a kernel discretization of the volume coupled to a high-order matrix gradient estimator and a Riemann solver acting over the volume ``overlap.'' We implement and test a parallel, second-order version of the method with self-gravity \&\ cosmological integration, in the code {\small GIZMO}:$^{\ref{foot:gizmourl}}$ this maintains exact mass, energy and momentum conservation; exhibits superior angular momentum conservation compared to all other methods we study; does not require ``artificial diffusion'' terms; and allows the fluid elements to move with the flow so resolution is automatically adaptive. We consider a large suite of test problems, and find that on all problems the new methods appear competitive with moving-mesh schemes, with some advantages (particularly in angular momentum conservation), at the cost of enhanced noise. The new methods have many advantages vs.\ SPH: proper convergence, good capturing of fluid-mixing instabilities, dramatically reduced ``particle noise'' \&\ numerical viscosity, more accurate sub-sonic flow evolution, \&\ sharp shock-capturing. Advantages vs.\ non-moving meshes include: automatic adaptivity, dramatically reduced advection errors \&\ numerical overmixing, velocity-independent errors, accurate coupling to gravity, good angular momentum conservation and elimination of ``grid alignment'' effects. We can, for example, follow hundreds of orbits of gaseous disks, while AMR and SPH methods break down in a few orbits. However, fixed meshes minimize ``grid noise.'' These differences are important for a  range of astrophysical problems.

4: 

5: \end{abstract}

6: