rowland

A very small unit of length used by spectroscopists between 1887 and the first decade of the 20th century.

In 1866 the Swedish physicist Anders Jonas Ångström (1814-1874) published a groundbreaking series of observations of the solar spectrum, giving the wavelengths of about a thousand spectral lines.  In the decades that followed, measurements by other physicists of the same lines gave wavelengths slightly shorter than Ångström’s. Almost 20 years later, in 1885, Ångström’s colleague Robert Thalén published an explanation1 of the discrepancy.

When a standard of length is made from a prototype like the International Prototype of the Meter, it is not possible to make the standard exactly the same length as the prototype. Instead, the workers first manufacture a physical standard whose length (in this case, the distance between scratches on gold dots near the ends of a bar) is as close as they can come to the length of the prototype. They then measure the standard repeatedly, using the prototype as their reference. The average of these measurements becomes a correction factor that is supplied with the standard. 

Ångström’s measurements of wavelengths were all based upon a meter standard at the University of Uppsala. The correction factor for this standard was supplied by Tresca, a distinguished French metrologist. Ångström took the marked length to be 999.94 millimeters. In fact, it was more nearly 999.81 millimeters. This discrepancy created a systematic error in Ångström’s measurements of wavelength. His ruler was shorter than he thought, which made the values of his measurements higher than they should have been.

In 1887 the American physicist Henry Augustus Rowland (1848-1901) published new measurements of the wavelengths of lines of the solar spectrum2. Thanks to the finer diffraction gratings made by the ruling engine Rowland had invented, his data were at least ten times more accurate than Ångström’s, which at that time were still in widespread use. 

Ångström had expressed his results in multiples of 10−10 meter. If instead one interpreted Ångström’s values as expressed in multiples of 999.81/999.94 th of  10−10 meter, the earlier results became consistent with Rowland's more accurate data. This quantity (999.81/999.94th of 10−10 meter) began to be treated as a unit in dealing with the old data, and came to be called a rowland. With the defining of the international angstrom in 1907 and the accumulation of fresh observations, the need for the rowland dwindled, and by 1910 it had disappeared.

1. Robert Thalén.
Sur le Spectre du Fer Obtenu à l'Aide de l'Arc Électrique.
Uppsala, 1885.

2. Henry A. Rowland.
On the Relative Wave-lengths of the Lines of the Solar Spectrum.
Philosophical Magazine and Journal of Science, series 5, volume 23, number 142, page 257 (March 1887).

home | units index | search |  to email Sizes drawing of envelope |  acknowledgements | 
help | privacy | terms of use