January 2008 Newsletter, Vol. 37, No.1


Spectrum of Rain

Figure 1

Rain Spectrum
Of the dozens of visually observed spectra that were published in the 19th century, J. Rand Capron’s plate of the “rainband” spectrum is surely one of the most attractive. Initially it appeared only in monochrome in Symon’s Monthly Meteorological Magazine1 but was eventually reproduced in color by John Browning in the 2nd edition of his popular handbook, How to work with the Spectroscope.2 It shows Fraunhofer’s solar lines, A - F, with the so-called rainband just to the red side of the D lines, increasing in darkness from top to bottom. Nearly everyone knows that the pair of lines at D are due to sodium absorbing in the sun’s photosphere. Lines A and B are due to oxygen, C and F are caused by hydrogen. Line E is a mix of unresolved iron and calcium lines. For a while, reference was made to “Big A” or “Little b” and more letters were added. It wasn’t long before they gave up using letters. A solar atlas of 1966 lists about 24,000 lines.3 These six spectra are supposed to illustrate the appearance of the solar spectrum as viewed through atmospheres containing increasing amounts of water vapor. The caption for spectrum No. 6, for example, reads “Spectrum seen July 6th, 1881. Rain-band everywhere, and exceptionally strong, stretching nearly half way between C and D. Whole spectrum darkened and obscured.” Late in the 19th century, the rainband was the subject of a small frenzy of interest in the possibility of remote detection of water vapor in the atmosphere and the prediction of rain. It represents the earliest attempt to apply spectroscopy to atmospheric remote sensing and meteorology.

By the middle of the 1860s it had become understood that the fixed dark lines seen in the spectrum of the sun were due both to solar and atmospheric absorption. As early as 1833 David Brewster noticed that certain lines were darker when viewed toward the horizon compared to directly overhead and in 1845 W. A. Miller mentioned that “...a singular appearance accidentally presented itself to me the other day. I was examining the spectrum of the diffused daylight towards the evening when a violent thunder-shower came on; lines not before visible were distinctly apparent, and a group in the brightest part of the spectrum between D and E, though nearer to the former line, became very evident, increasing in distinctness with the violence of the shower; as the rain passed away they again faded and disappeared.”4 Jules Janssen confirmed Miller’s observation in 1864 and then traveled to the summit of the Faulhorn, in the Bernese Alps, where he convinced himself that these lines were substantially weaker when observed at high altitude. He then made observations across Lake Geneva, over a distance of 21 km, using a large wood fire as a continuum light source and found the lines to be darker and more pronounced when observed over water. Finally, in 1866, he confirmed the features due to water vapor by making observations in a 37 meter length of pipe filled with steam.5 Gladstone seems to have anticipated the value of these observations in his 1868 review of the subject in which he concludes “No one can tell what secrets lie hid in these atmospheric lines, but to us it seems that by their careful and systematic observation the ‘Message from the Stars’ which has taught us so much may be rivalled in practical importance by a ‘Message from the Sky.’”6

Figure 2
Royal Observatory Rain Band

Charles Piazzi Smyth, Astronomer Royal of Scotland, observed the water vapor bands in the red region of the sky spectrum as early as 1872 and eventually proposed, in the 22 July 1875 issue of Nature, that observations of these lines could be used to predict rainfall. While enduring a long period of rain in Paris he observed that a strong dark line appeared to the red side of the D line in the spectrum of the sky while his barometer was “uninfluenced in its serene height and steadiness.”7 Smyth, a quintessential Victorian scientist who deeply enjoyed the scrutinous pleasures of scientific inquiry, was particularly interested in communicating spectral information clearly in graphical form. His plots of rainband spectra are interesting early examples of attempts to represent intensity as a function of wavelength (in his case always given in British inches, a unit which he considered to be divinely inspired). Figure 2 shows a portion of one of his plates of rainband spectra, published in 1877.8 It is interesting to compare his attempt at graphical representation of the darkness of the lines with the shading used in Capron’s lithograph.

Figure 3

Grace's Spectroscope
The instrument maker John Browning soon issued several special pocket spectroscopes designed specifically to show the rainband and marketed them as useful tools for the prediction of rain. The top-of-the-line model shown here, the “Grace’s Spectroscope,” was issued in 1884 and sold for about $473 in today’s currency. His competitors, Adam Hilger, The London Spectroscope Co., S. C. Tisley & Co. and John J. Griffin & Sons, among others, also began to advertise rainband instruments. Although a following of devotees emerged, a great deal of controversy developed over the reliability of the method. During the late 1870s and throughout the 1880s, many letters were exchanged in Nature, the London Times, and various meteorological journals arguing on both sides of the issue, sometimes very entertainingly. Capron, one of Smyth’s most devoted followers, published his “Plea for the Rainband” in 1881 and then, feeling that the issue was settled in his favor, “The Rainband Vindicated” in 1885. With Browning’s exuberant marketing and the enthusiasm of Capron, Smyth and Cory, who published How to Foretell the Weather with the Pocket Spectroscope in 1884, interest was sustained through the end of the century. From this distance, I doubt that any of us can fully appreciate the genuine novelty of these pocket spectroscopic observations to the curious Victorians. The colorful solar spectrum with its several dozen delicate dark absorption lines makes a fascinating image and the idea that it communicates knowledge about the composition of the stars and the earth’s atmosphere was undoubtedly compelling. Browning even sold a small spectroscope to be worn as a charm, in gilt or nickeled finish.

Since the eye is a notoriously inaccurate estimator of light intensity and without any fixed scale by which to make comparisons, all such visual observations were highly subjective and unreliable. The fundamental idea of rainband spectroscopy was sound but the technology of the period was simply insufficient to the task. The history of rainband spectroscopy has been well summarized by Austin9 and Peterson.10

Today, the spectral features of water vapor are routinely used for the monitoring of atmospheric moisture. In fact, water vapor is an important greenhouse gas and its distribution and concentration in the atmosphere is critical to atmospheric models as well as to weather forecasting. The water vapor bands which Smyth and his colleagues observed were weak O-H stretching modes of the water molecule which have only in recent years been studied in detail.11 The band which he favored, near 592 nm (0.00002331 inches), actually consists of about 203 separate lines, part of the 5nH2O vibrational band. No longer limited to the human eye as a detector, we now use much stronger transitions in other regions of the spectrum.

– Ben Smith
Department of Chemistry
University of Florida
E-mail: bwsmith@ufl.edu

References Cited
1 Capron, J. Rand, “A Plea for the Rainband,” Symnon’s Monthly Meteorological Magazine, 16, 181-190, December, 1881.
2 Browning, J., How to Work with the Spectroscope, 2nd edition, W. J. Johnson, London, 1883.
3 Moore, C. E., Minnaert, M. G. J. and Houtgast, J., The Solar Spectrum 2935 Å to 8770 Å, NBS Monograph 61, Washington, 1966.
4 Miller, W. A., “Experiments and observations on some cases of lines in the prismatic spectrum produced by the passage of light through coloured vapours and gases, and from certain coloured flames,” Philosophical Magazine, III, 27, p. 85, August, 1845.
5 Janssen, M. J., “Sur le spectre de la vapeur d’eau,” Comptes Rendus, 63, 289-294 (1866).
6 Quoted by Capron, J. Rand, “The Rainband Vindicated,” Symon’s Monthly Meteorological Magazine, p. 130, October, 1885.
7 Piazzi Smyth, C., “Spectroscopic prévision of Rain with a High Barometer,” Nature, 12, 231-232, 1875.
8 Piazzi Smyth, C., Astronomical Observations made at The Royal Observatory, Edinburgh, XIV (1870-1877), Neill and Company, Edinburgh, 1877.
9 Austin, J., “A forgotten meteorological instrument - the rainband spectroscope,” Bull. Scientific Instrument Society, 1, 9-12, 1983.
10 Peterson, T. F., Jr., “The zealous marketing of rain-band spectroscopes,” Rittenhouse, 7:3, 91-96, May, 1993.
11 Carleer, M., et al., “The near infrared, visible and near ultraviolet overtone spectrum of water,” J. Chem. Phys., 111, 2444 - 2450, 1999.

Return to January 2008 Newsletter

Primary Navigation

Isis and Osiris, Current Bibliography, Isis Books Received, Newsletter, Executive Office Publications


History of Science Society

440 Geddes Hall
University of Notre Dame
Notre Dame, IN 46556

574.631.1533 Fax