In this talk, I will review some highlights of my
studies of star formation in the past 35 years.
I started my PhD thesis on the theory of the stellar IMF
in 1977 at MPE in Garching and completed it in 1981.
I studied two different models: (a) hierarchical
cloud fragmentation (star formation as a random
multiplicative process) and (b) competitive accretion
in a protostellar cluster. The first model predicted a
log-normal stellar mass distribution (down to substellar
masses) while the second model produced a power law
(with a slope x = -1, close to the Salpeter slope).
I will outline both models and discuss how they stood
the test of time.
Later, as a postdoc at ROE in Scotland (1983-87), I became
an observer (mostly at UKIRT) and turned to near-infrared
(J,H,K) observations of young embedded star clusters,
such as the Orion Trapezium Cluster, using infrared arrays.
We observed near-infrared stellar luminosity functions
and derived the corresponding stellar mass spectrum,
using time-dependent mass-luminosity relations based
on pre-Main Sequence evolutionary tracks (without accretion).
A key cluster we studied (with HST) in the near-IR was
R136/30Dor in the LMC, and we proved the existence of a
low-mass pre-Main Sequence population in this starburst cluster.
In the 1990s, we carried out the first direct imaging studies
of young low-mass pre-Main Sequence binary stars and also the
multiplicity of massive stars, using 2D speckle interferometry
and adaptive optics observations.
We also discovered the first molecular hydrogen (H2) jets
from deeply embedded low-mass protostars (HH211, HH212).
Finally, time permitting, I will describe how I turned from a
near-infrared stellar astronomer to an interstellar
far-infrared astronomer, working with the B747SP
air-borne Stratospheric Observatory for Infrared
Astronomy (SOFIA) at NASA-Ames for the last 6 years.