In 1909, shortly after he arrived, Schwarzschild was joined by a colleague from Göttingen, Ejnar Hertzsprung, who stayed in Potsdam until moving to join De-Sitter in Leiden in 1919.
The first color-magnitude diagrams (an early version of the present-day Hertzsprung-Russell diagrams) to be published appeared in 1911, plotting stars of the Pleiades and the Hyades. Vogel had earlier attempted to classify stellar spectra, coming up with a scheme similar to the better-known scheme by Secchi - Vogel believed that his classification represented successive stages in stellar evolution, from young white stars to old red ones (an excusable error).The adjacent image is the Color Magnitude Diagram for the Hyades with magnitudes on the horizontal axis (absolute on top and 'absolute photographic magnitude' below, and color index on the vertical axis. These axes are the other way round to the current Hertsprung-Russell Diagram.
In 1911, he found that Polaris was a pulsating variable (i.e. a Cepheid). This had previously been assigned a magnitude of 2 and other stars assigned values relative to this, so Hertzsprung thereby showed Polaris to be an unreliable standard.
In 1913 he developed the method of distance determination using Cepheid variables, and used this method in an attempt to estimate the distance to the Small Magellanic Cloud (this was the first distance determination of an object outside our galaxy). Henrietta Leavitt had originally derived the Cepheid luminosity/period law by studying Cepheids in the SMC and made the realtion known in 1912. These clouds were not recognized at the time to be satellite galaxies of the Milky Way, but the assumption that all the stars within them were at more or less the same distance from Earth was nevertheless a valid assumption to make now and then. Leavitt did not identify the said stars as Cepheids - she said that ‘they resemble the variables found in globular clusters, dimishing slowly in brightness, remaining near minimum for the greater part of time, and increasing very rapidly to a brief maximum’. Hertzsprung showed they were similar to Cepheids and made the first attempt to calibrate Leavitt’s relation, i.e. to introduce absolute magnitude. Hertzsprung's estimate to the Small Magellanic Cloud was seriously in error because of the then unknown effect of dust absorption. (Shortly afterwards, Shapley made use of this relationship in a big way to derive distances to several globular clusters).
He developed a technique for observing double stars, using the Great Refractor, which eliminated errors to the extent that results were ten times more accurate than before.