Using a model for refractory clouds, a novel algorithm for handling them, and the latest gas-phase molecular opacities, we have produced a new series of L and T dwarf spectral and atmosphere models as a function of gravity and metallicity, spanning the \teff range from 2200 K to 700 K. The correspondence with observed spectra and infrared colors for early- and mid-L dwarfs and for mid- to late-T dwarfs is good. We find that the width in infrared color-magnitude diagrams of both the T and L dwarf branches is naturally explained by reasonable variations in gravity and, therefore, that gravity is the "second parameter" of the L/T dwarf sequence. We investigate the dependence of theoretical dwarf spectra and color-magnitude diagrams upon various cloud properties, such as particle size and cloud spatial distribution. In the region of the L$\to$T transition, we find that no one cloud-particle-size and gravity combination can be made to fit all the observed data. Furthermore, we note that the new, lower solar oxygen abundances of Allende-Prieto, Lambert, & Asplund (2002) produce better fits to brown dwarf data than do the older values. Finally, we discuss various issues in cloud physics and modeling and speculate on how a better correspondence between theory and observation in the problematic L$\to$T transition region might be achieved.

accepted to the Astrophysical Journal, 21 figures (20 in color); spectral models in electronic form available at http://zenith.as.arizona.edu/~burrows