Adriaan Van Maanen - Known For Faulty Results

Known For Faulty Results

In the first and second decade of the 20th century there was a large controversy in the astronomical world about the size of the Milky Way, the size of the universe and the nature of spiral galaxies. The result of this controversery was the great debate in Washington DC between Harlow Shapley and Heber Doust Curtis in April 1920. Shapley believed that spiral galaxies weren't systems equal to the Milky Way and in one of his arguments he used the results of Van Maanen, who measured the rotation of the spiral arms of spiral galaxies. After extrapolating his results he concluded that motion of the arms of spiral galaxies had a period of 10^5 years. While Curtis had stated that spiral galaxies were alike the Milky Way and thus had to have a similar size, believed to be 5 kpc at that time, the period of the motion was far too small for a radius of 2.5 kpc. At that period and radius, it would mean that the spiral arms rotated with a speed greater than the speed of light. Curtis agreed that if the results of van Maanen were correct, Shapley was right. But Curtis rejected the results of van Maanen due to the claimed (and unrealistic) accuracy of van Maanen. Later astronomers have re-examined the measurements of Van Maanen, and concluded he had made a serious error.

The origin of van Maanen's measurement errors emerged from his stereo blink comparator, which van Maanen used to compare new plates with plates some 10–20 years old. By blinking between those two plates he could detect small discrepancies on the positions of the objects. His reference objects were the field stars fairly near the edge of the plates. However he didn't consider that due to optical effects these stars had been smeared out a little towards the edges, slightly different for the two plates. This caused systematic errors resulting in imaginary movements.

Another possible explanation is that van Maanen simply saw what he had been trained to see for years. The belief that the "spiral nebulae" were relatively nearby and therefore ought to have a quite detectable rotation was quite widespread in the early 1900s and therefore very hard to ignore. This does not, however, explain supposedly corroborating findings from Mount Wilson, Lowell Observatory, Russia, and the Netherlands.