"Ideological crests versus empirical troughs: John Herschel's and William Radcliffe Birt's research on atmospheric waves,

Vladimir Jankovic
Program in the History and Pilosophy of Science
309 O'Shaughnessy Hall
University of Notre Dame
Notre Dame, IN 46556


The year was 1843, and the theme of English meteorology was measurement. Sir 'Thunder-and-Lightning' William Snow Harris was given his last British Association grant to complete the Plymouth series of over 120 000 thermometric observations, publication of which proved a costly venture, not least because the series implied no meteorological theory whatsoever.

In July of that year, however, John Herschel wrote to William Radcliffe Birt that the atmosphere might be considered 'a vehicle for wave like movement which may embrace in their single swell & fall a whole quadrant of a globe.' The idea of 'atmospheric wave,' thought Herschel, might well make sense of the odd series of London barometric readings made in September 1836, but more significantly, it might also lead towards solving the notorious 'storm controversy' of the 1830s between the American meteorologists William Redfield and James Pollard Espy. If Birt would accept, Herschel would propose him to the British Association for the Advancement of Science as the director of the new project to discover laws of weather behavior.

Birt had already acted as Herschel's computing assistant in the arrangement and reduction of many series of barometric measurements and was enthusiastic to accept the offer. The British Association eventually entrusted him to investigate the properties of atmospheric waves, but in doing so, however, it unwittingly prepared the ground for the failure of what Joseph Agassi once called Herschel's 'philosophy of success.' After five reports for the British Association and several contributions to the Philosophical Magazine, Birt dropped the research in 1849 without a conclusive explanation of mid-latitude atmospheric disturbances. He declared that the six-year effort was less than rewarding and subsequently returned to astronomy, his first scientific love.

Some time ago, Bruno Latour and Steve Woolgar suggested that scientists 'invest their credibility where it is likely to be most rewarding,' that is, where their intellectual capital immediately acquires a high value due to the supply and demand of the scientific 'market.' They add that a scientist's assessment of the fluctuations of the market explains (his or her) reference to 'interesting problems' and moves between 'problem areas,' in which a scientist enters into collaborative projects, grasping and dropping hypotheses as the circumstances demand. I will not here be much concerned with economic models of scientists' professional behavior, but the metaphor of credibility-investment seems fruitful means to make sense of that which motivated the research on atmospheric waves undertaken by Herschel and Birt during 1840s. The state of the sciences in 1840s, as Cannon has observed, was characterized with 'unexplained phenomena which were powerful enough to compel some sort of theories, and scientists were available adequate to supply theories.'

Indeed, Birt's and Herschel's publications and correspondence reveal the complex of interests that in the early years of the British Association created an increased demand for meteorological research among the members of the British scientific clerisy. Schemes for gathering observations were drawn up and, when the rhetoric of method changed to emphasize the value of hypothesis in science, theoretical concepts were advanced to explain data. Herschel may thus be said to have invested his intellectual credibility in 'atmospheric waves' when he claimed them a candidate for the explanation of large atmospheric disturbances. For the reasons examined below the idea proved increasingly difficult to present as a hard scientific currency. Its explanatory value progressively decreased to eventually render it hard to maintain and impossible to sell: in 1850, the Royal Society did not think it worth a 50 government grant.

In particular I will propose that the Birt-Herschel research emerged as a response to contemporary rhetoric of the puerility of British meteorology and the efforts of the British Association for the Advancement of Science in the way of ameliorating the condition. The failure however of the research to make an impact on the mainline theory of storms may be explained not only by reference to the theoretical problems that plagued the research from the beginning, but also as a result of the changed public legitimation of meteorology which in the early 1850s sought practical rather than theoretical results.

Even though the prominence of the Humboldtian sciences during early years of the British Association has often been stressed, meteorology received a relatively meager treatment both within and without that general agenda. This lacuna is unjustified in the light of historical evidence. In the Preliminary Discourse (1830) Herschel, the pater familias of the contemporary British men of science, described meteorology as 'one of the most complicated but important branches of science' in which 'any person who will attend to plain rules' might do effectual service. With Charles Babbage he tinkered with barometers and took his turn in the Grand-Tour 'barometrizing' in the Italian Alps; in 1833, on the way to the Cape of Good Hope, he incidentally re-discovered the equatorial depression in atmospheric pressure. In South-Africa Herschel campaigned to establish a network of observers; he published on matters of instrumental accuracy, measurement error, and reduction procedures. In 1836 he gave instructions for registering horary observations at the solstices and equinoxes, hoping to see them become 'great meteorological festivals when every observer must be at his post, over all the world.' It was due to this cosmopolitan enthusiasm that he began proselytizing for an extensive empirical enterprise directed toward understanding non-periodical barometric fluctuations or, what a German contemporary called, of the mutual reaction between the greater atmospheric currents over large geographical regions. The great unperiodic gales ('cyclones,' or 'storms' in modern vocabulary) were on that view considered to depend on causes extrinsic to the regular barometric oscillations, i.e., on those of 'local and transient nature' which an organized weather record would unmistakably bring to the light. 'A working society,' enthused Herschel, 'is wanted to go into all details and dwell upon the subject till something comes out.'

Put in context, Herschel's proposals for organizing the recording and reduction of observations, his plea for standardization and his later active role in promoting meteorological theory were reflections of a general discontent that the men of science had been for some time expressing in respect to the status of meteorology. While this discontent may be seen as a side-effect of the notorious debate on the decline of English science, it is more important to emphasize that meteorology suffered critical blows from much earlier period. In particular, the prediction of weather was a traditional target of criticism and doubt. In 1763, the Reverend William Borlase announced an impossibility of both understanding and foretelling the atmospheric changes 'from analogy and review.' In 1807, Louis Cotte asserted that '[the physics of meteors] is the least advanced part of meteorology. The useless attempts so far to find a theory to predict the weather from existing series of observations - these futile attempts, I say, prove that we do not know well enough the history of atmospheric phenomena,' In 1839, Henry Ince said that 'this dubious and very limited anticipation of the changes, which occur in the medium we breathe, is merely the fruit of personal assiduity and application,...it receives, therefore, no accession in the progress of ages, but perishes with the individual.' Whether fairly or not, the early nineteenth century commentators erupted with criticisms of a general lethargy that supposely prevailed in the investigation of weather-systems, of the insufficiency and profusion of observations, of the public uselessness of the existing stock of facts, and of the imprecision of means for standardizing and using meteorological instruments. It would not be off the mark to suggest that the theme of a history of the late Hanoverian meteorology would be the lament over its absence.

For instance, the meteorological committee appointed in 1814 by the Royal Society was so careless in overseeing the accuracy of instruments and the reduction of measurements, it led the Quaker cloud-classifier, Luke Howard, to suggest with acerbity that '[i]f this learned and highly respectable body feels the subject of the weather no longer worthy of notice, would it not be better, at once to dismiss the register from its transactions?' Worse yet, the register was the only part of the Philosophical Transactions acknowledged by the Council of the Royal Society, all other meteorological contributions being officially disavowed by the body. John Frederick Daniell, arguably the leading meteorologist of the day, added in 1823 that the weather record was so infamously inaccurate its publication on the continent would amount to 'the almost national disgrace.' To help improve the status of British meteorology, a group of metropolitan doctors led by George Birkbeck gathered the same year at the Ludgate Coffee House and founded the Meteorological Society of London. Daniell, Howard, and the eccentric polymath Thomas Ignatius Maria Forster were among Council Members, but their prestige and eagerness did not matter much as the society had its last meeting as early as in May 1824. Howard had moved to Yorkshire; Birkbeck had become busy with the London Mechanics' Institution. When the society was revived in 1836, it was taken over by astrometeorologists and continued its activities only unitl 1843.

In fact the Association had claimed exclusive rights on meteorological research. When William Harcourt and William Whewell, two visionaries of a provincially based meteorological research, asked young James Forbes for the first official British Association Report on meteorology, he spoke for the entire scientific establishment when lobbying for 'a total revision' of the 'infant science of Meteorology [which resembled] patches of cultivation upon spots chosen without discrimination and treated on no common principle.' Whewell wrote to Roderick Murchison that meteorology required 'multiple and extensive fagging.' Harcourt envisaged the average Association member as becoming, eventually, a useful weather-reporter and Herschel agreed: revamping was a pressing problem especially as the world was not 'yet civilized for meteorology.' More to the point, there was a sense that meteorology itself might become a civilizing force as the tenets of the early meteorological agenda, suiting Association's Anglican constituency of provincial 'soldiers of science,' managed to secure the ideals of social utility, collaboration, and in particular meteorology's socio-political benignity.

It is not surprising that in the context of the Association's enlightened conservatism meteorology's niche in the classification of sciences acquired more luster than in recent past. Earlier in the century, meteorology was classified as subspecies of either natural history, chemical philosophy, or physiology (meaning general inquiry into the natural world); now it slipped into the section I of the BAAS map of knowledge - the section of general physics. This elevated post was underscored by James Forbes's hopes to see the scattered empirical work straighten up by a species of mathematical physics modeled on Fourier's work on propagation of heat. Roderick Murchison and Edward Sabine went so far as to assess David Brewster's observations in Scotland as of the highest value to meteorology, 'the most interesting department of physics.'

Yet even the most interesting departments are sometimes plagued by prosaic snags; in the case of the early Victorian science of weather the two most prominent were the fiscal and physical inability to process already existing observations. Herschel continuously relayed his doubts about the Association's solvency in paying observers, computers and printers. Along with David Brewster, he thought scientific mediocrity lay in a misapplied Baconianism: 'perfect spontaneous freedom of thought is the essence of scientific progress,' and a society proposed along Harcourt's lines will only make science 'a democratic tyranny.' Meteorological observations made in the decade after this pronouncement brought the point home: 'witness the piles of unreduced meteorological observations which load our shelves and archives,' lamented Herschel, 'what we now want is thought, steadily directed to single objects, with a determination to eschew the besetting evil of our age - the temptation to squander and dilute upon a thousand different lines of inquiry.' So great was the concern at a runaway 'factology,' that Herschel advised South African observers to look 'with some complacency to their discussion and [be] induced to theorize and combine,' or, several years later, to narrow the field of inquiry, and concentrate on a distinct point in meteorology, 'such as tracing atmospheric waves.'

2. The storm controversy

Herschel's suggestion came at a time of great discussion taking place in America of the dynamic principles of mid-latitude and tropical storms; perhaps the most investigated episode of 19th century meteorology, the 'storm controversy,' besides being a scientific exchange of considerable weight, brought the importance of meteorological research to the public eye more than any of the utilitarian prefaces of the meteorological publications of the period.

In 1801 James Capper, a Cardiff-based East India Company retiree and amateur meteorologist, claimed that his experience with the Caribbean storms convinced him that 'a water-spout [was] only a whirlwind on a small scale, and a hurricane a whirlwind of an immense scale.' The 'rotatory theory of storms,' as the idea came to be elaborated during 1830s, held that the velocity of wind was a function of rotation of a vortex of fluid combined with its progressive motion. The theory was soon interpreted as the property of William Redfield, a New York transportation businessman and rising star of American meteorology, who in the paper on the North American and Atlantic storms (1831) argued with considerable empirical evidence that the mid-latitude storms blow in circles counter-clockwise around the center that advances in the direction of the prevalent wind. In the same year, the havoc wrought by a hurricane on the island of Barbados induced William Reid, a lieutenant colonel of the Royal Engineers responsible for rebuilding damaged government buildings, to spend his leisure time contemplating kinematics of storms. By 1838, Reid realized that Redfield's earlier results coincided with his own, and instantly wrote to John Herschel in hope that Redfield would soon 'have the great satisfaction of seeing [his] labours successful in proving the true nature of storms.' In August that year, Reid appeared at the British Association meeting in Newcastle and opened John Herschel's section with a paper expository of Redfield's 'Law of Storms.' The reception was positive and Redfield's and Reid's conclusions began to play a more prominent role among British weather scientists.

In North America, however, Redfield's position was far from self-evident. James Pollard Espy, Redfield's most outspoken opponent was at the time director of the Joint Committee on Meteorology of the American Philosophical Society. Espy insisted on thermal causes of storms: on his view air was violently sucked towards the storm's center, raised, and its vapory content condensed so that the released latent heat enhanced convection, which in turn accelerated centripetal movement of the air at the surface. Whereas Espy's ideas fell on fertile soil within the French Academy of Science, the theory remained a nonstarter in Britain. Alexander Bache did best to promote it among the British in 1836, but his efforts had no effect; in fact, John Dalton's reaction was plain negative. Nor was the subsequent visit of Joseph Henry more helpful; the meteorological doyen John Daniell, as Henry's diary shows, had already made his judgment in favor of the rotatory theory. Herschel inclined in the same direction: at the Newcastle meeting, the Reid/Redfield alliance owed its triumph in part to Herschel's own criticism of Espy's model. If the air was rushing toward the center of storm, wondered Herschel, why did the center exhibit the lowest atmospheric pressure?

Espy was particularly troubled with this objection. In 1839 he pleaded for Herschel's caution in judgment, 'for such is the weight of his name, that many will think it not worth while to examine a system which has been condemned by Sir John Herschel.' To amend the misunderstanding, Espy thought a personal appearance at the Glasgow meeting of BAAS in 1840 might help. He gave two talks but neither was received without considerable reservations. Espy acknowledged the criticisms and respectfully dissented.

While a consensus favoring 'whirling' over centripetal kinematics was gaining strength, dynamic issues were tackled with more diffidence. Redfield repeatedly invoked evidential support and argued for quo modo inquiry rather than rash speculation on veræ causæ. The only known cause of tornadoes and larger storms lay in his view in the tendency of all fluids to run in circuits when subject to opposing forces: 'all aerial fluids and fluviate bodies abound with the incipient rudiments of nuclei of gyration, in various stages of development, and these rudimentary gyrations seem to be the necessary results of variable and unequal corpuscular motion.' But to say that the 'gyration' could be a cause of 'gyratory' motion in storms was to argue in circle, some protested. Espy's thermal processes seemed a good alternative. Or was it electricity?

3. Surges of optimism

Herschel's understanding at this juncture was that the true theoretical candidate was not yet in the game. It was at this time that he and Birt, his computer between 1839 and 1843, began discussing alternatives, as the correspondence reveals. Birt had previously published in Astronomical Society's Notices on the periodical variations of b-Lyræ and a-Cassiopeiæ. Living near Bethnal Green in London, in 1831 and 1832 he was engaged in systematic astronomical observation; to John Lubbock he communicated a dozen of corrections to the celestial map prepared for the Society for the Diffusion of Useful Knowledge and even projected a survey of the Milky Way. Some time later, for Loudon's natural history Journal he wrote on the popular topic of cloud-formation and had a debut in the British Association Report with the summary of Wilhelm Dove's account of the aerial currents of the temperate zone. Herschel seemed to like Birt's arithmetical thoroughness and after returning from the Cape employed him on the project of reduction of barometric observations. From this time on, Birt primary scientific interests lay in meteorology.

In the early 1840s, Birt and Herschel worked jointly on the reduction, tabulation and graphical representation of barometric data; the search for diurnal and monthly regularities was the priority as Herschel hoped to lay down the empirical laws of the atmosphere and perhaps extend them into a comprehensive meteorological theory. But he also seemed particularly uneasy about the marginal role British gentlemen of science had in the storm controversy. The concern over a meager 'expenditure of thought' explains his outcry made at the Newcastle meeting (1838) and his call for more theoretical speculation of any kind; if nothing, it would serve as a temporary ferment for a potential controversy that 'might produce brilliant results, by the very collision of intellect.' Herschel's own contribution at this time was meant as an illustration of the plea, and was indirectly addressed to Reid, the only Englishman partly involved in the debate: perhaps the spots on the sun were tornadoes of the solar atmosphere and, if he were allowed to extend the analogy, would it be possible to imagine the solar atmosphere's own trade winds?

This was not what Herschel and Birt discussed in private, however; being immersed in a time-consuming numerology of raw data while witnessing the burgeoning if unsettled theories on the other side of Atlantic, they were about to end the 'plodding drudgery' of British-bred inductivism and to end it fast. While inquiry into the daily oscillation of atmospheric pressure had enthused many for more than half-century, including Snow Harris, Caldecott, Sykes, Kamtz, and Arago, Herschel came forward with a concept that would not only make sense out of what had already been tabulated but that would also had implication for the storm controversy itself:

Collingwood, July 28, 1843

Dear Sir,

....I have already come to the conclusion that our conception of a barometric wave can be easily enlarged (i.e. if we set out with my notion of their analogy to sea-waves) so as to take in vast (?) flattened (blanket-like if I may be allowed such a homely [comparison]) heavings more like the tide waves of the ocean - one such wave of at least 1000 miles in breadth is very distinctly traced over London in the September of 1836, at a rate of 21 miles per hour from NW to SE, and perfectly well made out... And if we consider (which is already well-known of the phenomena of the barometer) that we have often a rising [glass] for whole [month] and vice versa - and sometimes a general high or low level [...] for one or two months, it will be clear that our atmosphere is a vehicle for a wavelike movements which may embrace in their single swell & fall a whole quadrant of a globe. (my italics).

[...] My view of the matter to which I have alluded is that every barometric movement is essentially and intimately connected with an appropriate and corresponding wind -- as [...] every wavelike movement in a fluid consists of two distinct things - an advancing form and a molecular movement - [...] the advancing form is indicated by the barometer - the molecular movement by the wind, and between these two phenomena there exists of necessity a close and purely dynamical connection. [Knowing barometrical fluctuations one would be able to predict wind] for this could lead to prediction of great storms.

Later that year he repeated in press that it would be 'no small meteorological discovery' if by the study of barometric fluctuations one could formulate a law which would predict them, 'as by this we might possibly be led to the prediction of great storms [meaning small-scale thunderstorms].' As for the hot topic, 'the revolving gales' or 'revolving storms,' Herschel tred cautiously: he was not 'fully prepared to speak' on the subject, 'but some of the principal of their phenomena would seem capable of explanation in this way of conceiving winds of oscillation [those produced locally and non-periodically], and in which they would become traced up, not to 'funnel-shaped revolving depressions' [Reid-Reidfiled's hypothesis] in the nature of waterspouts, but simply to the crossing of two large long waves moving in different directions.' [See figure 1].

Figure 1

On this proposal, the possibility of explaining and forecasting revolving storms hinged on an accurate identification of the time and place of waves' intersection, an idea which was to remove all the theoretical entanglements developed in the Redfield-Espy dialogue. Empirically, Herschel thought the barometric record obtained at British magnetic and meteorological observatories already sufficed for a global wave dynamics. He himself would be unable to undertake this 'most interesting inquiry,' but was authorized by 'Mr. Birt to state, that should it be the pleasure of the British Association to entrust the subject to his inquiries by appointing him a committee for that purpose, he is prepared to do it if provided with a moderate grant to clear unavoidable expenses.'
'I am glad the investigation will proceed under the auspices of the Association,' Birt replied in a letter, 'and I will use every exertion to render the tracing of these waves as complete as possible.' In the first relation on atmospheric waves which appeared in the Association's annual Report Birt publicly returned gratitude to his mentor's recommendations. He thanked Herschel for providing Russian and European observations and George Bidell Airy for those from Greenwich. The rest Birt collected on his own, reduced them to sea level and standard temperature, calculated their daily means, transformed values into curves, and compared the results. Early curves confirmed the existence of the "beautifully symmetrical wave" occupying 13 days for its complete rise and fall. The evidence of its motion from Dublin to Munich promised 'a rich harvest of results.' Herschel was optimistic regarding Birt's methods of 'combining barometrical observations' which he thought was so successful as to prove 'we are dealing with realities with forms of a higher order than those which have hitherto been subjected to meteorological observation,' forms he believed analogous to the elliptic orbits astronomers infer from the data of successive planetary positions. For now, however, it would be best to thoroughly examine 'one or two good cases' as they would be worthier than 'a hundred touched out.'


The case in point was the 'Great November Wave' which swept Europe between 11 and 25 November 1842. [See picture 2]. When he first saw it outlined in Birt's graphs, Herschel exulted in claiming never to have seen anything 'more scientific' than the symmetrical curve as 'a simple result of direct observation. Such fact must have a meaning,' prophesied Herschel. It was a well defined specimen but more importantly, it seemed to be a recurring phenomenon, as Birt was gratified to inform Herschel in November 1845: 'I have much pleasure in acquainting you that I have great reason to believe the return of another extensive undulation [in November 1843].' In less than a month he wrote of a yet another recurrence, in November 1844, and in the Postcript to the Second Report written on November 27, 1845 Birt announced that the hopes to observe the Wave again that year were 'fully realized [as] the symmetrical wave has returned and has exhibited all its essential features.' Herschel was 'delighted': as mentioned, the 'periodical surges' of atmospheric pressure he saw as closely associated with and hence predictive of the 'revolving storms.'

Under the superintendence of the Honourable the Corporation of the Trinity House, London men of science made weather observations at what Royal Society had recommended as meteorological hours: 3am, 9am, 3pm, and 9pm. Birt decided that as the November barometric oscillations appeared to occur annually, observations should be undertaken on a larger scale, by interested individuals, keepers of the lighthouses, and on recommendation of Captain Beaufort, by the officials of the surveying vessels, all 'with a view to observe this particular wave.' Amateur gentlemen who would apply to Birt would be furnished with copies of instructions and forms for recording the observations. The journal would be intended for the period between October 1 and November 30, and would include readings of barometer, dry and wet bulb thermometer, the direction and force of the wind, and the weather at the times of observations.
Birt however soon realized that 'private enthusiasm and present body of observers' would not meet the requirements necessary for discovery of the global wave-system. Despite a generous financial support that the British Association provided for several past years, he came to believe that the time had come for a more active involvement on the part of authorities. Writing to Herschel in June 1847 about the lack of data for the identification of South-westerly movements, he wrote:

It has struck me that by the recommendation of the British Association Her Majesty's Government might be induced to direct that on a certain occasion hereafter to be determined upon observations should be made by the officers of H.M.Ships at certain of our foreign stations... The expense attending such a series of observations would be comparatively trifling as we may presume Her Majesty's Vessels are furnished with Barometers and the observations may be made and recorded four times each day by the officers of the watch.

Birt sent essentially the same message to Lieut-Col. Edward Sabine, 'the artful [scientific] dodger' able to handle 'everything and everybody.' . He hoped to enlist him for the wave cause believing Sabine could ensure participation of the Royal Navy in making barometric observations on ships and in naval stations (at that time Sabine was the general secretary of the British Association and foreign secretary of the Royal Society with important ties with the Admiralty). In July 1847, Birt also addressed Dr. John Lee, an eminent patron of astronomy and meteorology. In the long letter he presented a chronic lack of data which thus far prevented identification of atmospheric waves and discovery of the nature of winds. In his view, the beneficiaries of such a research would be the Naval, Merchant, and Military Services, as well as the members of the learned societies and directors of observatories. They all would have access to the standardized observations regularly deposited in a central meteorological office and arranged under superintendence of a Keeper.

To shew that I am no overrating the advantages to be gained, we have only to contemplate the results which our Transatlantic Brethren are arriving at in reference to those vast gyrative movements known as revolving storms or hurricanes by means of the system of observations established at the Military Posts of the US as well as by the interests taken by the Nautical Men of that country.

Both Herschel and Lee were enthusiastic about the scheme. Herschel endorsed the idea of a central Government office headed by a scientific meteorologist and employing 'a number of salaried computers,' provided '1st, that the Nation can afford it, - 2nd, that provisions should be made for the reduction and discussion of the observations so collected, and the regular publication of the results.' Lee saw no difficulty in meeting the conditions: '[M]y reply to his [Herschel's] first provision - I will venture to say that the Nation can readily, and ought cheerfully to afford it.' He thought the funds might be supplied by curtailing one regiment of Cavalry and one of Infantry as well as by commissioning one frigate of the Navy less in the year: 'if one million of the annual expenditure of 20.831.077 now spent on the Army, Navy and Ordnance were spent on the Arts and Sciences the public would readily approve of the appropriation of the funds.'

Birt cited Herschel's and Lee's favorable opinions in a letter to the Marquis of Northampton, the President of the Royal Society, and asked him to consider the beneficial effects that a governmental support might have on meteorology if the 'emphatically cooperative system' of observations were instituted within the Naval and Military Services. He argued that the same problem, a dearth of observations, had previously impeded Col. Reid's inquiry into the nature of 'Rotatory Gales' and was now a major obstacle in the wave research too. However, the problem was now even more pressing as the 'atmospheric wave as a meteorological phenomenon is altogether of a higher order than the revolving hurricane and that a latter is an immediate consequence of the crossing of two large waves, the tornado raging in the point of intersection.'

During this period, Birt also preached for the wave cause from the pulpit of the Philosophical Magazine; he solicited attention of potential observers to the importance of observations coming from 'any locality,' and gave a brief account of what the symmetrical wave meant at his recent presentation at the Cambridge Association meeting. It is not known whether at this time anyone followed Birt's instructions and what the return of the forms was. It is known however that in December 1847 Birt learned that [George Eden] Lord Auckland had asked Herschel to edit an instruction manual for officers on surveying and exploring expeditions. Herschel had acted immediately and asked Charles Wheatstone for meteorological instructions of general kind and was now writing to ask if Birt were willing to write those on detecting atmospheric waves. Of course, the two should stay in touch to avoid duplication. Birt replied that he was honored by Auckland's request and would do everything in his power.

Birt's appeals to individuals in the government offices and scientific bodies were bringing the wave research in the public eye; the empirical basis of the project seemed ensured and the results would appear only as a matter of time. Theoretically, problems had not yet arisen; Birt's structural analysis of the Great Wave of 1842 based on anenometric and barometric record gave strong probability to Herschel's idea that the molecular movement (a component of any wavelike movement) must be dynamically connected with the wind. Indeed, the molecular movement was asserted to be strongest in the wave troughs and directed towards them from each side, as Edward Sabine's had already suggested in the case of data from Toronto. This would explain the observed reversal of wind direction at the stations below the trough or the crest of the moving wave. If such were the case, the wind patterns, perhaps the greatest point of contention in the storm debate, could be in principle associated with readings of atmospheric pressure and thus connected with transit of waves.

Birt appeared more assertive on this issue in the review of Redfield's article on the paths of East-American storms. 'It is not my intention in the present communication to [...] attempt to substantiate or refute either the one or the other of the rival theories, [but the investigations of Redfield, Espy, Loomis, Reid, and others] have brought the inquiry to that point at which it becomes essential to connect it with some kindred branch of science, ... to strike out a path for working energetically in surmounting obstacles' which retarded understanding of the origins of storms. John Herschel had suggested, writes Birt, 'that they [storms] may be produced by the crossing of two large atmospheric waves moving in different directions.' To extend our knowledge of Redfield's 'rotatory gales' and Espy's 'centripetal hurricane,' it would be necessary to add barometric readings to the observations on wind, as they would greatly help meteorologists detect the origin of a gale arising from the intersection of two waves. 'Data collected during the American storm of February 1842 strongly implied two waves passing over the country with the storm raging in their intersection. Should the entire barometric observations taken over the United States on that occasion [...] support the theory of atmospheric waves, I apprehend Sir John Herschel's suggestions will be partly realized."

Behind the customary euphemism of Victorian scientific adulation, Birt nurtured an even greater faith in the potential of the wave theory of atmosphere. The occasion to voice it came with the request for contribution to A Treatise on Atmospheric Phenomena, compiled by Edward Lowe, a halo-aficionado living near Nottingham. One has a sense of subdued rivalry in Birt's opening remark that Colonel Reid had been unjust in saying that it was to William Redfield that 'we are indebted for the first true explanation of the cause of the rise and fall of barometer.' To be sure, Redfield's account was path-breaking in the discussion on barometric oscillations associated with storms, but there was more to weather than one fashionable if unavoidable topic. The more extensive undulations - waves - 'were left unexplained either by Mr. Redfield or Colonel Reid' and it was presumably up to their less lionized colleague to decide whether there was a similar arrangement of winds associated with wave-passage as the one identified in gyrating storms. Birt's enthusiasm turned tentative proposals into certainties when he pronounced that in the face of evidence 'a rotation of a greater or less extent and force must result from their [waves'] intersection; there appears to be a close relation between the rectangularly posited currents and the rotatory gales.'

These pronouncements mark an apex of the optimism Birt and Herschel nurtured in respect to the wave research and its potential in the study of revolving storms. It should be pointed out that from the beginning both had presented the research as a primarily theoretical enterprise. In a general sense, it was developed in the context of and in response to the status-of-meteorology rhetorics: a large scale empirical enterprise was thought tantamount to discovering laws of atmosphere. More particularly, it promised to facilitate the solution of the storm controversy through an alternative conjecture of the wave-intersection. It is in this context that we should understand Birt's governmental lobbying: it was concerned mainly with organization of data collection with relatively little to promise in the way of immediate public benefits, such as weather forecasting. In the mid-nineteenth-century Britain, this alone would prove detrimental to a research that threatened to consume large funds and ask for a considerable administrative engagement. But as the next section endeavors to show, the problems emerged even within the theory itself and eventually worked to rapidly erode legitimacy of the project.

4. Mounting obstacles

The idea of atmospheric waves is so simple that it could develop and be used almost colloquially. After a violent storm in Cornwall on 1 and 2 December 1763, William Borlase, perhaps the most conscientious weather observer during English ancien regime, made the following comment in his Natural History of Cornwall: 'By this violent and sudden alteration in the Mercury it appears that the atmosphere at these times must have been greatly agitated and proceeded over the earth in vast waves; low, deep, and hollow, when the Mercury fell: and mountainous afterwards, which occasioned as sudden rise.' Herschel's original proposal from 1843 did not diverge much from this common-sensical image of the heaving atmosphere: the wave was a representation of the barometric variation in amplitude as a function of time, featuring such ad-hoc terminological inventions in description of barometric curves (i.e. waves) as 'sudden undulatory movements,' 'abnormal protuberances,' 'trepidations,' or 'bulges.' In one of his analogical fits Herschel suggested that the conception ought to be enlarged 'till it approaches in some degree, in the extent of its sweep, and the majestic regularity of its progress, to those of the tide waves in the ocean.' He seemed to be a bit unrealistic about the success of big theoretical mergers.

With Birt's second Association report presented at the Newcastle Meeting in 1845, the problems multiplied. Some barometric readings could be accounted for only as result of the superposition of two or more waves moving in different directions and having different characteristics of the slopes. Some anomalies could be accommodated in this way, but more waves meant more questions about their origins, extent, velocity and so on.

Instead, Birt's strategy was to diminish a theoretical rift between Redfield's rotatory hypothesis and Dove's two-current theory. The latter presupposed storms as result of the interaction of two contrary currents for which Birt had already shown to be associated with the wave phenomena; Dove's theory was to link phenomena of gyrating storms with the 'wave-crossing' hypothesis. He thus associated anemonal data with the two-current picture and argued that the difference between rotatory gales and currents was in the 'first being a cylindrical body of air revolving around a central axis, while in the [latter] currents, or moving masses, are disposed longitudinally, and not circularly.' Both hypotheses would however predict similar phenomena: as the trough of atmospheric pressure approached an observer the velocity of the wind would increase, and after passage of the trough the wind would rapidly shift in direction. Thus in the third report for the British Association, Birt turned to comment on the observation of William Brown who in 1846 had proposed to explain barometric oscillations as result of 'the meeting of opposite or nearly opposite aerial currents.' Birt took some of Brown's data and in the discussion of them decided to 'adopt the wave hypothesis and ... regard the progress of barometric and anemonal phaenomena as the progress of waves.'

At this juncture, however, Birt seems to have realized that the concept of atmospheric wave might in the end fail to refer.

[T]he term atmospheric wave has been used to designate ideal individuality which the mind attributes to the process which it observes of the successive change of place which the barometric maxima and minima undergo [...] We readily see that the wave is a convenient method of representing the barometric fluctuations [and the barometric curve] does not represent the form of any reality in nature.

Birt's struggle to disentangle causes of minute irregularities on the wave-curve converted him into an awkward instrumentalist: the 'ideal individuality' was certainly a far cry from Herschel's Borlaseian naiveté. In addition, Herschel's idea to explain storms as result of wave-crossing was automatically reduced to the level of a mere model. On the empirical side, Birt found that the great symmetrical wave was to a certain extent erratic as to the period of its transit; the crest typically passed in mid-November, but in some years it occurred in October, on others in late November. As the initial and terminal minima of the observed readings were discovered to differ in no predictable manner, Birt warned that 'great difficulty must exist in the observation of the wave. If, for instance, any particular curve is expected to return, the observer in most cases will be disappointed,' especially if situated away from the locality of greatest symmetry. This qualification put in question usefulness of the proposed 'wave-manuals.' But it made them obsolete when he came to claim that even though a similar set of aerial currents recurred at the same time of the year, '[the great November wave] has no real existence: there is not a distribution of pressure over Europe at all similar to it.' The initial identification of the Great November Wave was due to a short series of observation; now, it was apparent that the distribution of atmospheric pressure was the result of two sets of parallel currents.

Of course, there was a great deal of tinkering in the way of securing the original idea. Recurrence of the waves indicated seasonal disposition of the atmosphere and could perhaps be used to probe into the features of Dove's currents. But nothing of the kind was executed. There was also a lingering hope of achieving a breakthrough in the storm controversy. In the Herschel-designed Admiralty Manual Birt proposed that the observations should be taken as 'capable of throwing light, not only on the most important desiderata as connected with storms, but also their connection or non-connection with atmospheric waves.' He used Redfield's observations on the occasional occurrences of the unusually high atmospheric pressure prior to the storm to point out that the event had to be associated with the wave-crest 'rolling forward.' But as for the question of why the crest would precede the storm, or what could cause the wave itself, Birt could not give any specific answer.

The time had come for several major disclaimers. We are fortunate to have two versions of Birt's perception of the shortcomings of the wave concept. The earlier one appears in the essay for Lowe's collection. He urged the reader to properly apprehend 'the precise idea attached to the term wave.' While the 'wave motion' and 'the wave form' might be conveniently represented by waves, it was

[n]ot so when we treat of the molecular motions or wind [originally Herschel's suggestion]: this is so entirely opposed to all our notions of molecular motions of waves, that we are called upon to hesitate before we designate the two ... a wave, merely because they exhibit, when contemplated as a whole, some wave phenomena.

In a letter to Herschel, Birt wrote that,

the principal objection to this definition of an atmospheric wave consists I apprehend in its non-agreement with hydrodynamical principles. So far as the term wave has been employed by myself in immediate reference to these particular phenomena I most readily admit it was ill-chosen and fails to convey a clear notion of such phenomena to the mind.

No wonder Birt directed his subsequent efforts toward more down-to-earth subjects; The Hurricane Guide (1850), derived from Colonel Reid's theory, was designed as a set of directions intended for seamen in order to avoid the centres of storms. A page long review in the Philosophical Magazine's gave a lukewarm praise to 'this little work,' and delivered the final blow to the concept: 'The fourth section,' wrote the reviewer (not unlikely to have been Herschel himself), 'speaks of periods of extra observations, and particularly of that of the great atmospheric wave of November. Our experience does not confirm the existence of this particular wave.'

Besides empirical and conceptual difficulties, in the late 1840s Birt also faced both professional and personal problems. In this period he searched for a post at various places (at the Observatory at the Cape of Good Hope, in Madras, and at the Kew Observatory). He also engaged in the researches of atmospheric electricity: at the Swansea meeting of the Association in 1848 he was requested to undertake the reduction and discussion of the electrical observation made at Kew, and the results formed a report of nearly ninety pages in length which was published in the Association's Report for 1849. In May 1848 he heard about the possibility of the British Association discontinuing maintenance of the Kew. He immediately asked Sabine to consider it a hasty decision in light of the five-year series of observations already made under the supervision of Francis Ronalds, the Honorary Director of the Observatory. If the Association was able to secure a specific grant for him, Birt himself would be willing to continue the work. In 1849, The Committee of the Association voted for the continuance of the observations and Birt moved on the premises of the observatory on November 2, 1849. His duties were initially limited to the discussion of electrical measurements, but at Ronalds's request he also commenced daily observations. He soon found himself 'become a mere machine doing work at such epochs which recording instruments are specially designed to accomplish.' Early morning observations and personal dislike that soon developed between him and Ronalds led to a nervous breakdown Birt suffered on 3 June 1850. On June 5, he jotted down a brief note to Sabine saying he was unwilling to undertake the duties of the Observatory another twelve months under present arrangements, these arrangements are not satisfactory to me nor are they such as I expected them to have been made.

Birt learned with consternation that the note was interpreted by the Magnetical and Meteorological Committee as his official resignation from the 'electrical' post at Kew. He immediately wrote an explanatory letter to the Visiting Committee. He was surprised, he wrote, that the information in a private note to an individual (Sabine) should have been so inconsiderately disclosed and misinterpreted by the members of the Committee. His reference to the termination of duties he explained as referring only to the early morning observations and not to the post as a whole. Birt's fight came to no good end however; after failing to obtain interview with Sabine and after receiving Association's terminating resolution he decided to pull out. The falling out between him and the Association was further exacerbated by the Association claiming to have rights on his most recent report on luminous meteors which he submitted after the involuntary termination. In the most dispirited moment, Birt wrote to Herschel in June 1850 to inform him of his complete retirement from scientific work whatever and to thank him for ten years of assistance and kindness.

Notwithstanding circumstances, Birt wished to continue investigation on atmospheric waves. In 1849 he asked for Herschel's and Phillips's advice on whether William Wheatstone's proposal to apply for recently instituted Government Grant of the Royal Society was justifiable in light of 'the progress I have been enabled to make under the auspices of the British Association.' Both approved the motion and Birt submitted a proposal which entered The Royal Society Council Minutes as request 'for the allotment of a sum of 50 pounds to enable him [Birt] to prosecute his inquiries into the nature and progress of atmospheric waves.' The resolution of this and other proposals was postponed to the session on 16 May 1850, when it a three-member committee was appointed to assess the allotment of the Government Grant for any specific scientific purpose. The body came to be known as The Committee of Recommendations for the Application of the Government Grants; Sabine was elected its Secretary. Whether because of the obvious bias in favor of Fellows or because the proposal was thought 'too theoretical or too philosophical to be 'good science'', on June 25 the Committee voted against funding Birt's proposal. In November 1851 he repeated the application and again was declined.

Birt's last meteorological work was the Handbook of the Law of Storms, a digest of the storm researches intended for use on ship-board 'without the trouble of consulting logs, or arranging isolated records of the wind, weather, etc.' The book may be regarded as a testimonial to the utilitarian concerns increasingly informing meteorological investigation. Any inclusion of theory seemed to have been in need of apology:

Impressed with the idea that 'abstract science,' under all circumstances, bears remotely, if not immediately, on the well-being of society, we have not shrunk from placing before the public our views of the storm paths in all localities, even at the risk of giving to our otherwise practical volume the appearance of a scientific volume, and incurring the oft-repeated cry, when immediate benefit does not appear, of 'Cui bono? '

In the section devoted to atmospheric waves Birt accordingly eschewed the 'wave-intersection' hypothesis and discussed waves as 'bearing somewhat on storms' to the extent that their troughs were now defined as rectilinear atmospheric streams of the hurricane strength. Prediction of those would be of considerable importance to seamen, but '[a]t present so little is known of the nature of atmospheric waves on the open ocean, that it would be premature to offer any further remarks on them.'

Birt's lasting scientific contribution was in selenography, to which he turned after the hapless accident with the Association's managerial machine. Despite that episode, in the period between 1865-9, the Association published five of his reports on mapping the surface of the Moon. His other writings dealt with the lunar activity, identification and catalogizations of the smaller features on the moon's surface. Three years before his death, in 1878, he was elected the first president of the Selenographical Society.

5. Conclusion

The meteorology of atmospheric waves received relatively meager attention among contemporaries and among historians of meteorology. In 1855 Urban Leverrier, the Director of Imperial Observatory in Paris, reported the results on movement of the atmospheric wave associated with the great storm that had in November 1854 fallen upon the French fleet stationed in the Black Sea. E. Liais, Leverrier's assistant, concluded that the direction and speed of the wave had no relationship to the direction and speed of its accompanying winds - a fact Birt had already shown to be contradictory to Herschel's hypothesis on the relationship. Leverrier hoped A.L.Cauchy would undertake mathematical analysis of the wave movements but there is no evidence it was ever done. Neither was Birt's research widely recognized. Henry Piddington, a storm theorist known for the invention of the term 'cyclone,' noted in 1860 that 'Sir J. Herschel proposed the idea, that two or more [...] extensive atmospheric undulations, or barometric waves, from traversing in different directions, intersect each other, and from their opposing forces cause the phenomena of hurricanes or rotatory storms.' In this century, Isaac Cline, the Principal Meteorologist of the US Weather Bureau, wrote as if Birt's 'wave period' never took place; referring to 'The Hurricane Guide,' Cline wrote that 'W. Radcliffe Birt, as early as 1852 [at which time the research was practically over for two years], showed unusually complete grasp of the movement of winds in cyclones.'

If it was Birt who in his late contributions to the BAAS Reports had already cast a shadow on the validity of the wave research, it was to Robert FitzRoy to produce the most scathing criticism of it. 'The first full time government appointed meteorologist,' FitzRoy claimed that 'what are commonly called 'atmospheric waves' are delusive; and that, although there are waves in any line indicating oscillations of the barometer, there are not such movements in the atmosphere itself, as are usually adverted to by expressions 'trough' and 'crest.''

The organization of meteorological research in Britain during 1850s emerged in public forums as a more pressing issue than Birt's theoretical investigation. In 1850 The British Meteorological Society was founded and was instrumental in mobilizing a number of voluntary weather observers. By 1851 William Reid influenced John Fox Burgoyne, the inspector-general of fortifications, to authorize the setting up of additional meteorological stations in colonies and in Britain; Burgoyne went forward with the idea of U.S. and British cooperation, and due to initial disagreements about the observational standards an international conference on meteorology took place in Brussels in 1853 and was a great success. In Britain, in August 1854 Robert FitzRoy was nominated President of the newly formed Meteorological Department of the Board of Trade. But as Jim Burton has observed, practical, not theoretical issues took priority in those proceedings: they addressed the safety of sea transport, profitability of trading which depended on reliable wind and sea-current maps, and the prospect of military advantages. 'Last was the interest of men of science who favoured the study of meteorology for its own sake.' In this situation, Herschel's initial uncompromising prodding for theory proved archaic as the new meteorological 'commodities' began to attract attention of the governmental authorities and scientists alike. His project to explain atmospheric disturbances appeared increasingly misplaced and even irrelevant.

In addition, as the wave episode suggests, the scientific 'patriotism' that in the ebullient theoretical market of the storm debate initiated Herschel's hopes of success and opened pathways for Birt to explore, in the end failed to carry through a daunting empirical cargo, the associated conceptual problems, and only secondarily through the strains between Birt and Association. In an ironic turn of events, the theory which Herschel's reputation and British Association's financial solvency helped promote to the forefront of English meteorology, lost in the face of phenomena it was contrived to account. FitzRoy made a shrewd retrospective judgment that '[g]reat authorities then countenanced the theory and apparently sanctioned such opinions. Yet there is so much argument against these views, that even the highest names may scarcely warrant their implicit adoption.' Nature, however loosely one may want to associate it with business of observation, appeared to have had the upper hand in a contention with ideological powers that be.