by Sarah Dry
The instrument made it possible to do solar physics at a moment’s warning, anywhere the fervent observer happened to be, liberating him or her from surveillance. Skill was required, a delicate habit of minute adjustments, to get good results. The instrument should be aimed low down near the horizon so as to subtend as thick a portion of the atmosphere as possible. A break in the clouds was ideal, but the sun should not interfere too directly, lest it overpower the delicate observation. Sunrise and sunset were also not recommended for this reason. Fogs, mist, and dense coal smoke could also obscure the readings.
For Piazzi Smyth, all this adjusting was enjoyable, part of the fun of looking at the skies. The convenient little tool found its way to his eye many times a day. Fifty times was easily possible when he had his “enthusiasm-fit” upon him.31 Every situation was different, every cloud break, every confluence of pressure, temperature, and wind bearing variously on the transformations of the weather. He didn’t know exactly what, if anything, he was looking for. He didn’t really need to be looking for anything, at any particular moment. He looked first, almost before thinking.
In 1875, Piazzi Smyth made a trip to Paris, where he visited the Astronomer Royal Urbain Leverrier (he found him exceedingly rude, leaving him and Jessie to find their way home alone through a thunderstorm). That storm followed the Piazzi Smyths back to London and, observing it carefully with his spectroscope, Piazzi Smyth noticed a hazy and indistinct—but nevertheless present—“dark broad band” crossing the spectrum between the red and orange sections. The band was darker than the areas surrounding it, and was fuzzy rather than distinct. It faded as he moved the spectroscope to another part of the horizon. When he took his eye away from the instrument and looked with plain sight at the sky, he could see nothing in particular that looked different in that region than in any other. Piazzi Smyth continued to observe as he traveled north to York, where he noticed the strong presence of a blurry dark band in the spectroscope one sunny morning when rain seemed unlikely. Rain did indeed ensue, and observing it fall, Piazzi Smyth felt vindicated in his hunch that the pocket spectroscope was a special new kind of tool for predicting the weather.
He set out to share the news of what he’d come to call the “rainband.” Unlike the hard-edged lines of the solar spectrum, which were understood to be caused by the absorption of light waves by different substances in the solar atmosphere, the rainband was blurry, indistinct, and dynamic. Just what caused it, Piazzi Smyth did not say, though his identification of it with the coming of rain was a strong indication that he thought it most likely to be the signature of water vapor in the atmosphere. He announced his discovery in a letter to the journal Nature, founded just months earlier, under the title “Spectroscopic prevision of rain with a high barometer.” The title made it clear that Piazzi Smyth was modifying what had previously stood as the most basic of meteorological assumptions—that a high barometer, and therefore high pressure, implied good weather.
FIG. 3.10. Solar spectra recorded under different weather conditions showing the position of the rainband at r.
In fact, Piazzi Smyth had revealed precious little about the rainband. He did not offer an opinion on what caused it, and did not even mention the phrase water vapor in his article. More than a new bit of scientific knowledge, what Piazzi Smyth was so eager to announce was a new tool for doing science. Spectroscopy—recently identified as a tool for diagnosing the contents of the earth’s atmosphere—could also be a device for measuring the changing amounts of those elements in the earth’s atmosphere. This made it a tool of practical meteorology, what he called the “prediction of weather for the common purposes of life.”
Despite Piazzi Smyth’s excitement, there was reason to be very cautious about the possibility that the “rainband” spectroscope might transform weather prediction. The blurry bands of the so-called rainband were even harder to learn to see than the fixed solar lines, since they were variable. They changed because the thing they represented—water vapor in the atmosphere—was itself constantly shifting. The allegedly convenient and easy-to-use rainband spectroscope was in fact an instrument for diagnosing variability of a very complex and very particular sort. While spectra contained an enormous amount of information about the entire breadth of the atmosphere at a glance, they were also snapshots, representing just an instant in time. To be useful, they had to be interpreted in relation to snapshots that came before and those that came later. Even the darkness of the bands was only useful when compared with preceding rainbands. Comparing the intensity of successive rainbands was an intensely subjective task, which only relatively few could master sufficiently to render it a consistent practice. F. W. Cory, writing in the Quarterly Journal of the Royal Meteorological Society, argued that the spectroscope was simply too difficult to use, requiring two or three months of “patience and perseverance” (this admitted by one of its most ardent supporters) to be mastered easily by unsupervised individuals.32
Piazzi Smyth persevered. He wrote a series of letters to popular journals in which he tried to reframe the problem of prediction within meteorology, a problem that had always been knotty but had become particularly so ever since FitzRoy’s sensational forecasts and their cancellation following his sensational death. For Piazzi Smyth, prognostication was not a bad word, and the pocket spectroscope no better or worse than existing methods for predicting the weather. It was merely one more tool for energetic, self-motivated people who couldn’t wait for a “perfect” meteorology (akin to astronomy) to rise from the ashes of Admiral FitzRoy’s storm warning project. “We need not after all be offended at the mere name of ‘prognostic,’” Piazzi Smyth reassured the public, pointing out that even the stalwart barometer could only go so far in predicting the coming weather. The boundary between folk wisdom and science, which the Royal Society committee was desperate to police, was a mere illusion, according to Piazzi Smyth. Knowledge was knowledge, however it was gained, and in relation to as complex a thing as the weather remained necessarily provisional, dependent on skill, judgment, and the individual perspective of a single observer whether it was generated by the most rustic fisherman or the most well-trained astronomical observer. “For are there not prognostics and prognostics in meteorology! What are not the risings and fallings of the wind-compelling barometer itself, but a weather prognostic for those who can interpret them.”33
In the face of the almost unimaginable complexity of the weather, Piazzi Smyth was both pragmatic and optimistic. Rather than limiting the sources of data or its applications, he thought it made sense to increase both. He imagined legions of independent observers across the country, “many, very many people,” the natural acuity of whose eyes was enhanced by the pocket rainband spectroscope to render them an army of proto-Supermen, able to see through the clearest of skies to the vapor that lurked within. Thus outfitted, they could “observe and speculate on the weather for themselves at their own places of abode, supplementary to any forecasts that may be issued once a day from London.” The best part about the spectroscope was its portability and ease of use. It enabled any individual to penetrate his solitude and immobility to reach, quite literally, beyond the confines of terrestrial domesticity and into the vastness of space itself. A glance of just two seconds was enough to “tell an experienced observer the general condition of the whole atmosphere.”34 It gave a feeling of security, Piazzi Smyth explained, to know that even in the most cramped and confined space, “with no more than a few cubic feet of peculiar, and for science-purposes, vitiated, air about it,” he could nonetheless still be “nobly looking through the whole atmosphere from the surface of the earth right through to space outside, and analyzing its condition as to watery vapour (the raw material of rain, as the Times phrased it) in one instantaneous, integrating glance.”35
The information the spectroscope gave to the observer was not merely fast but all-encompassing. It penetrated the entire atmosphere at the speed of light and made a single human being the di
agnostician of the globe. This was a global science of the atmosphere practiced by individuals. Unsupervised, untethered, unabashedly personal—the pocket rainband spectroscope fulfilled Piazzi Smyth’s fantasy of how science should be. Here was a tool with practical benefit to the sailors, farmers, and holidaymakers of Britain and beyond, a tool which could deliver immediately on its promise, rather than coyly holding out the hope for future understanding of the laws which drove the weather. With it, every man could become an observatory, twitching the skies above.
But what Piazzi Smyth considered the very best qualities of the spectroscope—the way it facilitated a multitude of quick and calibrating glances, its constant availability to those who found themselves gripped by it—proved to be its undoing. Instead of bringing meteorological observing to the people, as Piazzi Smyth had hoped it would, the miniaturized, portable rainband spectroscope revealed just how ill suited most people were to the practice of science. The craze for rainband spectroscopy faded as quickly as a summer storm, leaving the public to rely on their own eyes, their familiar barometers, and the unfamiliar new weather maps when it came to making decisions about the weather.
While Piazzi Smyth was promoting the benefits of individualized vision, the scientific winds were blowing the other way. A Parliamentary investigation into the Met Office was begun in October 1876. In the wake of FitzRoy’s death, the Met Office had embraced so-called self-registering instruments that could automatically trace the changes in the weather. These fantastical devices were like instrumental chimeras, combining the normally separate actions of measurement and registration in one object. The bedeviling influence of friction, which had foiled attempts to design self-registering instruments in the past, met its match in the form of photography. One of the first applications of photography was to the challenge of automatically registering the weather. In 1845, just six years after Louis Daguerre had pioneered the photographic process, two meteorologists (Francis Ronalds at Kew Observatory and Charles Brooke at Greenwich) set about designing a series of self-registering instruments (including a magnetometer, electrometer, barometer, and thermometer) which could deflect a beam of light against a photographic place. Other self-registering instruments used the simpler method of connecting an inked pen to the measurement device.36
These traces were to be used not for weather forecasts—which FitzRoy’s death had revealed as dangerously subjective—but for the long-term project of deducing the physical laws that underpinned the movements of the atmosphere. In the eyes of Robert Scott, FitzRoy’s replacement as head of the Met Office, the values of comparison and continuity completely trumped those of independence and skill. In 1875, he quoted approvingly from the 1840 report of the Committee of Physics and Meteorology of the Royal Society: “Systematic cooperation is the essential point to which at present everything else must be sacrificed; and cooperation on almost any plan would most certainly be followed by more beneficial results than any number of independent observations, however perfect they might be in themselves.” The contributions of those who the committee referred to as “amateurs of science” were welcomed only as long as they conformed to the rules, “even,” it was noted, “at the temporary sacrifice of their own views and convenience.”37
Exactly how the ever more numerous traces of pressure, temperature, and other weather-related phenomena could be transformed into a science of the weather was an open question. British meteorologists in the 1870s felt themselves to be stalled in a stage of early development. What astronomers had once achieved—the ability to make predictions that were accurate far into the future—seemed an ever-receding goal. Meanwhile, astronomy itself had forfeited the self-confidence it had assumed in the wake of Newton’s great achievement. Men such as William Herschel, Edward Sabine, John Herschel, David Brewster, Jules Janssen and, last but not least, Charles Piazzi Smyth had shown that astronomy could be a physical as well as a positional science, but in so doing they had exposed new sources of ignorance and uncertainty. The mature science of astronomy was young again, while meteorology, ever the “infant” science, sought new sources of confidence.
The scandal of FitzRoy’s death did not help matters in Britain. The promise of automatic observatories was a decidedly mixed one—offering the chance of fulfilling Humboldt’s dream of disaggregating the signals of a pluripotent Cosmos but challenging the ability of scientists to manage ever greater amounts of data. Data had long threatened to overwhelm singular astronomers as they faithfully recorded the skies in a single location, night after night. Once armies of self-registering instruments were unloosed on the observatories of the globe, it was hard to see how it would ever be possible to catch up.
New techniques for reducing the traces of such instruments were urgently required. Humboldt had understood this back in the 1830s and had urged Heinrich Berghaus to publish a graphical companion to his Cosmos in the form of a Physikalischer Atlas, which used diagrams to represent the way climate, plants, animals, and geological features changed across the globe. In Britain, Francis Galton came up with a strikingly visual way of finding the mean values for meteorological traces that involved superimposing a series of traces and graphically determining the average line. As innovative as these visual methods were, they could only go so far in the absence of data. In exchange for the personal judgments of forecasters, self-registering instruments promised objective knowledge and produced reams of data. How to wring meaningful understanding from the proliferating traces of the atmosphere was less than obvious. Not all data was created equal. One could have simultaneously too much data of one kind and not enough of another.38
One problem was that the atmosphere was three-dimensional but observers had been largely limited to data from the surface of the earth. This was part of the reason that Piazzi Smyth had been so taken with the spectroscope. It enabled anyone who used it to soar high overhead, traversing unimaginable distances in the process. This was its great advantage, but it was also a disadvantage, as the spectroscope was unable to distinguish the absorptive capacities of different parts of the atmosphere. Every molecule that lay in the line of vision was included in its gaze. It flattened the heterogeneity of the atmosphere even as it provided a way to diagnose its changes. This was a paradox that Piazzi Smyth, for one, was willing to accept, considering the benefits of this kind of vision accumulated to outweigh the losses.
But there were many ways to see the skies. One of the challenges was to get up into the atmosphere and observe it in situ. From the 1850s onward, a series of daring and popular balloon journeys up into the atmosphere took place. These sensational flights were undertaken in the service both of meteorological knowledge and with the spirit of adventure that characterized polar expeditions. They were, as far as they went, immensely successful in raising the profile (literally) of meteorology and capturing the imagination of the public. But balloons were expensive, and the journeys dangerous. At best, they provided a single set of observations recording the conditions in one particular column of air over the course of several hours. Generating systematic knowledge out of these singular adventures would be almost impossible.
Another way to get up into the skies, figuratively, rather than literally, was to pay close attention to clouds. Clouds rode the air currents that determined weather and climate. Noting their sizes, their shapes, and their movements was a way to map the invisible ocean of air above. Clouds were like flags in the upper atmosphere, telling an observer which way the wind was blowing and in which direction, as well as giving an indication of how much water vapor was present in the air. If that were the case, then clouds could be a way to see deeper into the mysteries of the atmosphere. Rather than obscuring the heavens, or frustrating astronomers, clouds could reveal the patterns of atmospheric movement and the laws that drove air around the planet.39
The first step was to classify the clouds. When Luke Howard had introduced his pioneering cloud nomenclature at the beginning of the nineteenth century, he had helped regul
arize the study of clouds. What he had not been able to say with confidence was whether his tripartite system of clouds would apply throughout the world. Were clouds globally uniform, or were certain clouds only to be found in certain parts of the world? The century had almost ended before anyone had seriously attempted to answer this question. In 1885, an amateur meteorologist with deep pockets named Ralph Abercromby decided to try. He set out on a self-financed circumnavigation of the globe with the explicit aim of determining how universally applicable Howard’s system was. He determined that while the same type of cloud could signify different kinds of coming weather in different places, the essential cloud types were indeed universal.40
Abercromby described his clouds with words, but his discovery inspired the search for more methods of faithfully capturing clouds. If clouds were universal, they could unlock the mysteries not only of the local weather but of what it seemed increasingly reasonable to assume were planetary weather patterns. But in contrast to temperature, pressure, or even rainfall, clouds resisted instrumental registration—they belonged to that “extensive class of phenomena which cannot be recorded instrumentally, but of which it is necessary to take careful notice owing to their importance as indicating changes which are in progress in the atmosphere.” Clouds were almost impossible to observe with the kind of objectivity that instruments promised, but they were too important to ignore. “It is very difficult,” noted C. H. Ley in the preface to his father’s contribution to cloud classification, Cloudland, “to treat of a vague and complicated subject in any but a vague and complicated manner.”41 What was needed was an instrument that could register the clouds in the same way that the barometer registered pressure and the thermometer temperature—instantly, faithfully, and reliably. By the 1870s, exposure times were fast enough that the photographic camera presented itself as a potential solution to this problem.
