We have developed a technique of IR spectral classification in which w
e use K-band spectra (R similar to 1000) to derive the spectral types
and continuum veilings of young, late-type stars (similar to 1 Myr, >
GO). We show close agreement between the spectral types derived in thi
s manner and those obtained optically. We complement previous optical
spectroscopy with IR spectra of the most heavily embedded members of t
he young, embedded cluster L1495E. We critically analyze the translati
on between observable (spectral type, photometry) and theoretical (Tef
f, Lbol) parameters and use these data to construct an H-R diagram. We
find that the evolutionary tracks of D'Antona & Mazzitelli imply a co
eval population of II Myr and a plausible initial mass function (IMF).
However, these models may underestimate the masses of objects near an
d below the hydrogen burning limit. The models of Swenson produce impl
ausibly old ages and the models of Baraffe et al. yield somewhat old a
ges and an implausible IMF. We use infrared imaging to show that the s
pectroscopic sample for this cluster may be seriously incomplete below
-0.15 hi,. After applying a completeness correction to the IILIF deri
ved with the tracks of D'Antona & Mazzitelli, we find no evidence for
a turnover at low masses; the IMF appears roughly flat in logarithmic
mass units. Compared to the results of photometric studies of rho Oph
and NGC 2024, the IMF appears roughly invariant among star-forming env
ironments representing a 2 order of magnitude range in the density of
young stars. However, the detailed behavior of the IMF from low stella
r masses into the substellar regime will remain uncertain until (1) be
tter evolutionary tracks are available and (2) the sources in the phot
ometric completeness correction can be spectroscopically conarmed as l
ow-mass cluster members.