Traditionally, the strong hydrogen emission in T Tauri stars was interpreted in terms of mass loss [Lago 1979, DeCampli 1981, Hartmann et al. 1982, Natta et al. 1988, Hartmann & Kenyon 1990]. Hartmann et al. (1990) and Calvet, Hartmann & Hewett (1992) compute hydrogen line profiles for spherically symmetric wind models and for ``cone'' geometry inner disk winds respectively. The resulting line profiles explain the H Alpha observations for some T Tauri stars but have difficulties in explaining the shapes of higher Balmer lines. The conclusion from these authors is that the wind models considered, in general, do not provide satisfactory results for the Balmer lines.
Pedrosa (1996) computed H Alpha line profiles in the context of a
radially spherically symmetric isothermal wind model. The radiative transfer
treatment is very detailed, and makes use of the Sobolev approximation.
The broadening mechanisms are Doppler broadening and Stark broadening
and the model includes electron scattering. The statistical equilibrium
is computed for a hydrogen atom with 15 main levels and all
its sub-levels plus continuum. The computed H Alpha line profiles predominantly
display a P Cygni shape with a deep absorption component. For high
densities (eg. a density at the base of the wind of 
particles
per cubic cm) the lines become stronger and broader with the
blueshifted absorption almost disappearing. The peak of the model line
profiles is always redshifted relative to the rest wavelength. As
pointed out by the author the model has a problem generating line
profiles broad enough without increasing the peak intensity beyond the
observed values, a problem that might be overcome by dropping the
isothermal character of the models. Some of the model line profiles
exhibit emission beyond the blueshifted absorption. When this happens,
the line tends to be symmetric at the level of the wings, with these
extending up to velocities higher than the maximum outflow
velocity. This is explained by Stark broadening of the line.
A different approach for wind models is taken by Grinin & Mitskevich
(1991) and Mitskevich, Natta & Grinin (1993). The
assumption of a continuous medium for the wind model is dropped and a
clumpy structure is assumed. The stellar wind consists of separate
clumps of material which decelerate in the outer regions of the
flow. These so called ``stochastic wind models'' find that for filling
factors of about 10% the model line
profiles for H Alpha are double peaked and similar to many H Alpha profiles
observed in T Tauri stars. For even lower filling factors and wind mass
losses smaller than 
the lines
become more symmetric. The main problems of the computed H Alpha line
profiles when compared with the observations are: the smaller separation
between emission peaks, by about 100 km/s and, as pointed out by Reipurth,
Pedrosa & Lago (1996), the excessive sharpness of the blue
side of the main peak.
Bertout (1977), Bertout (1979) and Bastian (1982) compute Balmer line profiles for infalling material, investigating the formation of these profiles in YY Orionis stars. The models considered are either spherically symmetric, in Bertout (1977) and in Bastian (1982), or axisymmetric, in Bertout (1979), with radial infall. Although model profiles are IPC as observed, the detailed shape of the lines are not reproduced.
Calvet & Hartmann (1992) abandoned the assumption used in their previous
papers that the hydrogen lines in T Tauri stars are formed in a wind and
investigated the possibility that Balmer lines arise in infalling
envelopes. Using a ``cone'' geometry, as a very rough approximation to
magnetospheric accretion, they compute line profiles of the lines
mentioned above. Their results show that H Alpha can display little
evidence for infall while lines higher
up in the Balmer series have redshifted absorptions. These authors
suggest that Balmer line emission originates in an accretion column
with a wind being responsible for blueshifted absorption features but
not contributing substantially to the emission. Hartmann, Hewett &
Calvet (1994) explore the formation of the hydrogen Balmer lines
in an accretion
scenario further, by taking a more realistic geometry for the
infall. The flow is now considered to follow the dipolar magnetic field
lines from a region near the inner edge of the accretion disk onto the
stellar surface. The Sobolev approximation is used throughout in their
computation of line profiles and a two level
hydrogen atom is used. The line profiles computed by these authors are
centrally peaked and slightly blueshifted. The presence of redshifted
absorptions depends on the inclination angle of the system and on the
temperature of the accretion shock. For large inclinations one sees the
accretion column against the hot region on the stellar surface where the
accretion shock occurs and redshifted absorption features tend to appear
in the higher members of the Balmer lines. H Alpha rarely displays
redshifted absorption. At lower inclination angles the accretion column
is seen against the cooler stellar photosphere and no absorption is
produced. The only hint for infall is the more extended red wing in the
line profiles. Muzerolle, Calvet & Hartmann (1998) extend the
previous work to a multilevel hydrogen atom in statistical equilibrium. The
calculations presented show a good match between line fluxes and line
profiles for the Balmer lines and also between model line profiles of
Br Gamma and WL16's Br Gamma line profile presented in Najita, Carr
& Tokunaga (1996). Edwards et al. (1994) present high resolution
observations of Balmer line as well as Na D and HeI 
for a sample of 15 T Tauri stars. They find
widespread evidence for infall either by identifying redshifted
absorption features in the line profiles or by interpreting the more
symmetrical profiles based on the work by Hartmann, Hewett & Calvet (1994).
The hydrogen lines in T Tauri stars are interpreted to arise from recombination of ionized hydrogen [Natta et al. 1988, Martin 1996]. Both for wind and accretion models, the ionization mechanism is thought to be photoionization by Balmer continuum photons [Natta et al. 1988, Martin 1996] since outflowing/accreting gas is opaque to Lyman continuum photons.
A wind scenario for the formation of the hydrogen emission lines observed in T Tauri stars has been the most compelling during the years. The predominance of blueshifted absorptions in the H Alpha line profiles, the success in modeling some of these observations with wind models, together with the unequivocal presence of winds in T Tauri stars has been the main driving force behind this interpretation. In recent years there has been a shift towards an interpretation based on the magnetospheric accretion model, however the debate is not yet settled. Also, for a given star, both winds and accretion flows may contribute to hydrogen line emission, with different lines being formed in distinct regions and/or affected differently by those regions.
The successes and failures of the above models in explaining hydrogen line profiles in T Tauri stars have been based on the study of lines from the Balmer series. The sole exception is the work by Najita, Carr & Tokunaga (1996) where the Br Gamma emission line is studied for a small number of objects.
The observational work on near infrared line profiles (Pa Beta and Br Gamma data) and the radiative transfer modeling of those lines, presented in the chapters that follow as well as testing the magnetospheric accretion model, aim at contributing for a better understanding of the origin of hydrogen line emission in T Tauri stars.