“It is a hobby of mine to have an exact knowledge of London. There is Mortimer’s, the tobacconist, the little newspaper shop, the Coburg branch of the City and Suburban Bank, the Vegetarian Restaurant, and McFarlane’s carriage-building depot.”Sherlock Holmes in ‘The Red-Headed League’ (1891), reading a street by its trades
In 1889 the Post Office London Directory listed, street by street and house by house, who lived there and what they did. This map reads it the way Holmes read that street: one circle for every square kilometre, its size the number of entries, its colour the trade that stands out there.
Doctors gather round Harley Street, barristers in the Temple, watchmakers in Clerkenwell, cabinet makers in Shoreditch, sailmakers by the river. The tour moves at random through such patterns, each with the evidence from the time that it was really so.
Click a circle for its trades. Scroll to zoom, drag to move.Tap a circle for its trades. Pinch to zoom, drag to move.
Part of the Holmes Hub. This map is one page of holmeshub.online, a free hobby project that puts the Sherlock Holmes stories on the London of 1889: every place Doyle names on the maps of the time, the city in 3D, and the stories themselves to read. The same Directory fills the houses of that 3D city: where a house of 1889 could be matched it shows who lived there, and where it could not, the house is roughly placed along its street by its number, and the card says so.
The tour. For each stop it shows the evidence that the pattern was really there: Charles Booth’s survey of London labour, the Survey of London, the guidebooks and pictures of the day, and a story or two found along the way.
Made. The street section of the Post Office London Directory for 1889 (“Kelly and Co., Printers, Gate Street, Lincoln’s Inn Fields”, says its title page; “Price forty shillings”; the scan is on archive.org, public domain) was read by OCR, column by column: street headings with lines. Each line is a house number and a name, usually followed by a trade. A street goes on the map on evidence, in this order. First its side streets: the Directory names them inside every street (“here is Poland street”), and where two or more of them cross a street of that name today, that is the street ( streets; checked on streets whose place was already known, half fall within 61 metres and nine in ten within 361). Then its postal district: the street of that name nearest the middle of the district, within the reach of that district ( streets). Last a name that occurs once in London, when no district could be read ( streets). And one rule works the other way: when the map of today knows two or more of the side streets and none of them touches the street, it is a different street of that name, and it stays off. Together: streets with lines, % of the whole. Each street sits at one point, and the points are gathered into squares of one kilometre; the squares with at least 150 lines are drawn. The trade is read from the words after the first comma, by 51 patterns (“watch” or “clock” is a watch & clock maker), first match wins.
Trust. The Directory lists householders, shops and firms, not lodgers or the poorest families, so the East End is thinner here than it was. About half of the lines are not on the map: streets whose heading the OCR could not read, streets renamed or gone since 1889, and the seven very long sections. A whole street counts at one point, so a square can borrow a few entries from next door. OCR misreads letters (“lodging honse”), and a line with two trades counts once. “Stands out” compares a square with all placed lines, not with all of London. Read the circles as a large, honest sample, not a census; each stop of the tour says where the pattern is confirmed by a second source.
“By each of these things a man’s calling is plainly revealed.” — A Study in Scarlet (1887)