Pannekoek began systematically observing the night sky and recording these observations while he was still in secondary school. Some of his observations of
variable stars later found their way into his scientific publications. In the winter of 1889–1890, he recorded the variability of
Polaris, which had been suspected before but was not confirmed until 1908 by
Ejnar Hertzsprung. He also tracked the variability of
Algol, which formed the foundation for his PhD thesis. Another early interest of Pannekoek was the
appearance of the Milky Way. In 1898, Pannekoek published a series of articles in which he articulated how he thought the Milky Way should be observed. Here, he explained that minor details should be tracked through verbal descriptions, while the general distribution of light should be tracked to
isophotic lines. Pannekoek published his own observations of the northern Milky Way in 1920 in the form of drawings, isophotic diagrams, and verbal descriptions. He also combined his observations with the independent observations of
Cornelis Easton,
Otto Boeddicker, and
Julius Schmidt to create a composite image of the Milky Way, which he called the "mean subjective image". A few years later, he also published his observations of the southern Milky Way, which he made during an eclipse expedition to the
Dutch Indies in 1926. He also developed a photographic method to represent the light distribution of the Milky Way. To capture the clouds that formed the Milky Way, Pannekoek deliberately recorded photographic plates out of focus, which caused the light of stars to spread into disks, allowing their light to overlap. These plates were then measured and combined into tables showing the surface brightness of the Milky Way. Pannekoek's research on the appearance of the Milky Way was closely connected to his research on the structure of the
galaxy. In this research, he adapted the statistical methods of
Jacobus Kapteyn to investigate individual clusters in the galaxy. His most important results were the measurement of the distance to the star clusters responsible for the Milky Way clouds in
Cygnus and
Aquila. He determined that these were located at a distance of 40000-60000
parsec from the sun. This was much more distant than was commonly assumed to be the diameter of the entire galactic system. This result provided early evidence for
Harlow Shapley's expanded galaxy. The final decades of his professional career Pannekoek mostly spent on researching the
astrophysics of
stellar atmospheres. In his theoretical research, Pannekoek explored ways to expand upon
Meghnad Saha's ionization formula to better understand the physical conditions in the outer layer of stars. One of his main conclusions was that the narrow spectral lines in c-type stars, as found by
Antonia Maury, were most likely caused by lower pressure in the stellar atmospheres. In the 1930s, he developed theoretical models for the atmospheres in order to reproduce the entire spectrum of a star, but failed to produce a model that was entirely satisfactory. In 1935,
Rupert Wildt showed that this was because Pannekoek had underestimated the impact of the
H− ion as a source of optical opacity. In observational astrophysics, Pannekoek produced the curve of growth for
Deneb in 1931, the first for a star other than the sun. He and his students also published comprehensive catalogues of the spectral lines in late type stars based on photographic plates taken by Pannekoek at the
Dominion Astrophysical Observatory in 1929. Because of these studies, he is considered to be the founder of
astrophysics as a separate discipline in the Netherlands. Pannekoek was also part of scientific expeditions to observe
solar eclipses in
Sumatra and
Lapland. He was also interested in the
history of astronomy and his book,
A History of Astronomy, is considered a standard reference on the subject. Pannekoek became member of the
Royal Netherlands Academy of Arts and Sciences in 1925. His work in galactic structure, astrophysics and the history of astronomy was of international renown and won him an honorary degree from
Harvard University in 1936, as well as the
Gold Medal of the Royal Astronomical Society in 1951. The crater
Pannekoek on the Moon and the asteroid
2378 Pannekoek are named after him. The
Anton Pannekoek Institute for Astronomy at the
University of Amsterdam, which he founded, was named after him in 1982. == Thought ==