Tools and methods

Press interest in airships, January-April 1913

'Everybody's Doing It' was the name of a popular revue which opened in the West End in February 1912; the music and lyrics (including a near-eponymous song) were co-written by Irving Berlin. It was also the Manchester Guardian's stab at a contemporary pop cultural reference to describe just how widespread the phantom airship scare had become by the start of March 1913. There are more concrete ways to express this than ragtime. Geography is one; chronology is another.

The graph above shows two things. (After relying on Plot for many years, I've switched to DataGraph, which is not free but is more powerful and much easier to use.) The blue bars represent the number of British periodicals (mostly daily newspapers, London and provincial) which mentioned mystery airships on each day in January-April 1913, while the red bars represent the number which mentioned airships, whether mysterious or non-mysterious (for example, the activities of German or British military dirigibles). It doesn't matter whether a newspaper mentioned scareships once as a humorous aside or devoted half a page to a topic, both are counted equally here. Three phases can immediately be distinguished. (I must admit to having fudged the data a little bit: I've assumed that every issue of the Aeroplane would have mentioned airships, as I don't have access to copies to check. Flight certainly did.) The first, from the start of January through the third week of February, is characterised by a relatively low level of press interest in airships, in which references to mystery airships predominate (though not so much towards the end of this period). The second phase is clearly the peak of the phantom airship scare, the last week of February and the first week of March, when more than two or three times the usual number of periodicals talked about airships, overwhelmingly the mysterious kind. The third phase extends from the second week of March until the end of April. There are far fewer mentions of scareships here, even compared to the first phase. But interestingly, the amount of attention paid to airships in general remains very high: several times that of the first period, and not too far short of that in the second, peak period.
...continue reading

Western horizon from London, 21 February 1913, 8.30pm

The planet Venus normally sticks close to the Sun and so can only be seen very shortly after sunset, to the west (or before sunrise, to the east, when it is a morning star). But every 584 days, when it reaches maximum elongation in its orbit, it is far enough from the Sun in apparent terms that it remains visible for quite some time after dusk. It also relatively close to the Earth at this time and so unusually bright: only the Moon is brighter. At such times Venus dominates the western sky and it can be very startling, especially for the infrequent stargazer.

As it happens, Venus reached maximum elongation on 11 February 1913, right in the middle of the phantom airship scare. The above thumbnail probably isn't very clear, but the full-size version, made with Stellarium, shows the western horizon from London at 8.30pm on 21 February 1913, the beginning of the scare's peak. (London without any buildings, light pollution or clouds, admittedly, but the view would have been roughly the same from anywhere in the British Isles.) Venus can be seen low above the horizon, almost exactly due west, and extremely bright (apparent magnitude -4.1, though extincted by the atmosphere to -3.2). Anyone who happened to glance in that direction would see a brilliant light hovering in the distance, very different to the other stars and even planets. If they watched it for a few minutes they might see it drifting northwards and perhaps sinking lower; if there were clouds scudding by or trees waving in the wind the effect might be enhanced. It would be very easy to think an aircraft was flying about, equipped with a searchlight.
...continue reading


View Scareships, 1913 in a larger map

Here's where the 1913 phantom airship sightings took place. Actually, there are a few from late 1912 (including the Sheerness incident), the blue ones. Red indicates sightings in January 1913, green February, cyan March, and yellow April.

A quick visual inspection shows that the density of sightings was greatest in Lancashire and Yorkshire, in a belt running from Liverpool in the west to Hull in the east. However, this perception is skewed somewhat by the cluster of reports from about a dozen places in and around Selby on 21 and 22 February. Clearly something happened there that night, and whatever it was can in no way be discounted, but Yorkshire would stand out far less otherwise (though there would still be the sightings from around Hull and Grimsby, including that from the City of Leeds, about the last of the whole scare to generate widespread interest in the press). By contrast, the clusters around Manchester and Liverpool, though smaller, are also more sustained, with reports spread out over four or five weeks. Other significant groupings include the south-east of England and, relatively late in the scare, the east coast of Scotland. And, of course, the coastline either side of the Bristol Channel, which featured prominently in the scare from almost the first to almost the last. Cardiff was the place most frequented by mystery airships, with four visits recorded at long intervals; that the Chief Constable of Glamorganshire was one of the first to see it and was willing to publicly appeal for witnesses to come forward may help explain why hundreds or even thousands of people saw them there. Only Hull presents a similar example of a mass sighting, though there were others where smaller crowds gathered to watch the scareship. It's also worth noting that there were occasional reports of mysterious airships from Ireland and from the Orkneys.
...continue reading


Having updated my list of online early 20th century British newspapers, I have mostly good news to report. The most exciting development comes from Wales. I have previously lamented the total lack of digitised Welsh newspapers from the period 1901-1950, and it appears that in large part the reason for this is that the National Library of Wales has been busy scanning and OCRing, and the first fruits of its labours are now available at the Welsh Newspapers Online site. Already there are 14 titles available, some in Welsh, some in English, some in both:

Aberdare Times
Aberystwyth Observer
Llangollen Advertiser and North Wales Journal
London Kelt
London Welshman
Pembrokeshire Herald and General Advertiser
Prestatyn Weekly
Seren Cymru
Swansea Gazette and Daily Shipping Register
Tarian Y Gweithiwr

Future digitisation plans are ambitious: I count 55 titles with 20th century content scheduled to added later this year, and there's more to come, including many journals. The OCR quality seems very high, which is crucial for search, and the article interface is really very nice and pleasant to use. Best of all, Welsh Newspapers Online is completely free. It's fantastic that Wales has decided to make its cultural heritage open to the world in this way; most other UK newspaper archives are locked up behind a paywall.
...continue reading

In May 1909, the three major organisations promoting aviation in Britain, the Aeronautical Society, the Aero Club, and the newly-formed Aerial League, announced that they would henceforth coordinate their efforts. The Aerial League would be recognised as 'the paramount body for patriotic movements and for education', the Aeronautical Society 'the paramount scientific authority on aeronautical matters', and the Aero Club 'the paramount body in all matters of sport, and the development of the art of aeronautics' (Flight, 8 May 1909, 258.) These are important organisations in the history of British aviation. I've visited the Aeronautical Society (now the Royal Aeronautical Society), to use their library (now part of the National Aerospace Library) and I've visited the Aerial League (now the Air League), to examine its archives, but I've never been to the Aero Club (now the Royal Aero Club), to see what it has. And now I don't have to; or at least soon I won't have to. (Though, actually, most of the material is in the RAF Museum London's collections, which I have visited for other reasons.)

Andrew Dawrant has left a comment on my post about Claude Grahame-White which brought to my attention the existence of the Royal Aero Club Collection. The Collection exists to preserve and promote the Aero Club's historical material, whether generated by itself or donated to it, including photographs and postcards, fine art, and trophies and other artefacts. (The Aviators' Certificates, i.e. pilot's licenses, which were awarded by the Aero Club are available through Ancestry, alas not for free.) But what really caught my eye is the digitisation programme. In the future this will include the Aero Club's papers (an index is already available). Moreover, the minutes of the Aero Club's executive committee from 1901 (i.e. the beginning) to 1956 have been scanned, OCRed and put online. Admittedly (as I know from looking at the equivalent Air League records) it is in the nature of minutes that they generally record only resolutions proposed, resolutions voted, letters read at the meeting, and not the cut and thrust of the discussion and debate. And as agendas were set in advance (and members no doubt wanted to get off home), they are often oddly silent on the great matters of the day, even when they would seem to be of direct relevance. But even so there is a tremendous amount of information to be gleaned from them, even just on a basic level of who knew who and did what when.

This is a great resource and I thank the Royal Aero Club for making it available and accessible to the public free of charge.

1 Comment

I've updated my list of online British newspaper archives. This time, the new titles are:

Aberdeen Journal
AJR Information
Catholic Herald
Connacht Sentinel
Cork Examiner
Jewish Chronicle
Irish Press
Irish Times
Kilkenny People
Louth and North Lincolnshire Advertiser
Nenagh News
Northants Evening Telegraph
The Post/Sunday Post (Dundee)
Sligo Champion
Sowerby Bridge News

Many of these are Irish, mostly from from the Irish Newspaper Archives but a couple from the British Newspaper Archive. [edit: I actually forgot my self-imposed rule to only include Irish newspapers before 1922 -- oops!] Thanks to a list of newspaper archives at Wikipedia, I also found the Irish Times, a couple of London-based Jewish newspapers (one, AJR Research, free) and the free Sowerby Bridge News, which appears to be an individual effort. Another great free resource is the Catholic Herald -- for example, here's their response to Guernica, 'Open town bombing a regular practice of Reds'. In addition, 14 titles have increased year coverage, though sometimes this is only a year or two.

With respect to the US-based NewspaperArchive, there's some good news, some other good news, and some bad news. The good news is that they have overhauled their interface and it's now much easier to use. For example, they have adopted Trove-style filters so you can quickly narrow a search by time period or location or title. This means it's more useful for discovery purposes, whereas before it was more difficult than it had to be to find things that you didn't already know were there. The other good news is that they have mostly cleaned up the metadata for their British titles. Previously if you went looking for British titles from the 20th century you would get a lot from the 18th especially, but also the 17th and 19th centuries. Maybe poor OCR was to blame (e.g. 1744 looking like 1944) but it's something that could easily have been noticed and fixed if a human had looked over things now and then. It seems that has now happened; while I did find a couple of errant titles the vast majority were correctly dated (or at least, the vast majority of the ones said to be from the 20th century were; it could be that the reverse problem is occurring too). So these are really positive developments, and signs that NewspaperArchive is committed to improving its service. The bad news, however, is really quite bad, at least from my point of view. It seems that NewspaperArchive has run into copyright or permissions problems, because now none of the British titles it carries go beyond 1904 (which also means that a couple of them, Lloyds News and the Staffordshire Sentinel, have been dropped altogether; the Hackney Express and Shoreditch Observer has been added, but this is already in the British Newspaper Archive anyway). Previously they mostly extended into the 1910s or 1920s. For example, the single most useful title, the Daily Mail, used to be covered up to 1921, so you'd get Versailles, the Great War, the People's Budget, the dreadnought race, the beginnings of airmindedness... now you get, I don't know, the Taff Vale decision? So I doubt I'll be renewing my subscription to NewspaperArchive: there's just not enough stuff for me now. But given the great improvements NewspaperArchive has made recently, I'd certainly join up again if the coverage situation improves in the future.


Uses of 'Mars' and 'canals' vs uses of 'Mars' only in peer-reviewed astronomical articles, 1861-1970

So, to wrap up this accidental series. To check whether professional astronomical journals displayed the same patterns in discussing 'Mars' and 'canals' as the more popular/amateur ones I again looked at the peak decade 1891-1900, this time selecting only the more serious, respected journals. However, because of the French problem I had to exclude L'Astronomie and Ciel et Terre (the former was apparently more popular anyway). So for my top three I ended up with Astronomische Nachrichten, Publications of the Astronomical Society of the Pacific (PASP) and Monthly Notices of the Royal Astronomical Society (MNRAS). Astronomische Nachrichten ('astronomical notes') was the leading astronomical journal of the 19th century, founded 1821. It published articles in a number of languages including English. Fulltext Service seems to be multilingual, as it picks up the German (at least) equivalents of Mars/Martian and canal/canals. That doesn't help with the French problem, but that will only affect a small minority of Astronomische Nachrichten's articles. The Astronomical Society of the Pacific was founded in California as a joint amateur-professional organisation. Its PASP is now a very highly regarded journal, although I must admit I don't know if this was always the case. MNRAS is the journal of the Royal Astronomical Society in Britain. It also happens to be where my solitary peer-reviewed astronomy article was published (and when I say 'my', I think approximately 1 sentence relates to research I actually undertook), but even so it really is a highly-respected journal.
...continue reading

In my post about the lingering scientific interest in the Martian canals hypothesis after 1909, I said that there was a problem with journal coverage. What do I mean by this? Have a look:

Uses of 'Mars' and 'canals' in peer-reviewed astronomical articles

This is a repeat of the first plot in the previous post, showing the number of articles published in peer-reviewed astronomical journals mentioning 'Mars' and 'canals' between 1861 and 1970, only this time for each of three journals: Observatory, Journal of the British Astronomical Association, and Popular Astronomy. I chose these three because they were the journals which had the most such articles, both over the entire period and in the peak decade of the 1890s.
...continue reading


Uses of 'Mars' and 'canals' in peer-reviewed astronomical articles

In a recent, hmm, let's call it a discussion resulting from an old post I wrote about the US Air Force's one-time interesting in mapping Mars, I tried to assess how scientific interest in the Martian canals hypothesis lingered after the early 20th century, and said I would run up some figures to illustrate the data. So here they are.

My source is the ADSLabs Fulltext Service. ADS is the Astrophysical Data System, an online database of articles published in astronomy and physics journals. Which doesn't sound so amazing these days, but it was in 1994 when I first used it! (More on its history here.) The interface has changed remarkably little since then, but it is still free and very comprehensive. While it is primarily an abstract service, fulltext is available for many older articles -- but only as non-searchable images. Moreover, not all articles have abstracts. However, the text of articles from most of the major journals have been OCRed into a parallel database, the Fulltext Service. Like the classic ADS Abstract Service, this was not designed with historians in mind, but it's still quite useful.
...continue reading


There's been a huge amount of interest on Twitter and in the media about the new Bomb Sight website, developed by the University of Portsmouth with assistance from the National Archives and elsewhere, and deservedly so because it's fairly excellent. In short it's an interactive map of the London Blitz compiled from a number of sources, showing where what kinds of bombs fell when. So you can browse to (or search for, though this has been temporarily disabled due to high traffic) your favourite part of London and see why there's a mid-20th century building interrupting that otherwise Victorian facade. Zooming in you see a marker for each bomb fall, with a link for more information. You can also get a statistical overview for each borough or ward, or for Greater London as a whole. Each location links in with relevant Blitz photographs sourced from the IWM, as well as related stories from the BBC's WW2 People's War site. In the map view, you can flip between the aggregate bomb census over seven months, or a single week's worth of bomb falls, or just the first twenty-four hours. You can also overlay the original Home Security maps from which the census data is derived, which is valuable because, thanks largely to the Blitz itself, some streets which existed in 1940-1 can no longer be found. As a bonus, London's invasion defences can also be displayed, using data taken from the Council for British Archaeology's Defence of Britain Dataset. An AR app is on its way, though sadly only for Android devices, not iOS.

Despite popular impressions, Bomb Sight doesn't show all the bombs recorded falling on London during the Blitz, but only those recorded between 7 October 1940 and 6 June 1941 as well as those recorded on 7 September 1940. That is, nearly the first month of the Blitz is missing. (This is quite clearly stated but it would be easy to overlook.) The reason for this is that the weekly bomb census maps only began to be compiled a month into the Blitz. It might be possible to fill in the gap from other sources; that's what has been done with the first day of the Blitz, which is from London Fire Brigade records via the Guardian. Of course, coming from different sources the data will be disjointed but that is inevitable with this kind of project. I gather that it is also intended that all the weekly censuses between October and June will be added to the site, which would mean you could slide through the weeks to see how the bombardment changed over time; or else further research might pin down the date of each bomb (at the moment when you click on one it only gives the census period, i.e. up to an eight month period). It would also be extremely interesting to compare bomb falls from the First World War, if only to illustrate the differences in the scale of bombing.

Also, Bomb Sight only covers London. Again, this is at least partly due to the nature of the data sources. But again it's something which could be remedied. Other, smaller Blitz maps like this have already been done for Southampton and for the West Riding of Yorkshire (for the night of 14 March 1941; from after the Blitz there is also Londonist's V-2 map. Depending on the permissions and formats, and hand waving wildly, it should be possible to aggregate these maps into Bomb Sight, if desired. Or else somebody else could build a website to do the aggregation. But in the end, somebody would probably have to do the hard work of sifting through local ARP records to generate the data for outside London, assuming those records do exist and are detailed enough.

None of which is intended to cavil at what the Bomb Sight project has achieved, as it's very good stuff indeed.