Archive: May 2017

  1. Superposition of Fort Lockhart view from Saragarhi

    Leave a Comment
    Superposition of Fort Lockhart view from Saragarhi
    By guest blogger Richard A. Fowell (heliography expert)

    The analysis showing locations visible to both forts places Saragarhi quite well in the East-West direction for a given position North or South, but is a bit fuzzy in the N/S direction. [1]

    The period photos taken of Fort Lockhart from Saragarhi, however, can be used to accurately place Saragarhi in the N/S direction, but will be weak in the East-West direction.

    By combining the two sources of information, we should be able to confirm the position of Saragarhi post without the need to lean on the statement that it was at the “highest” point on the ridge.

    There are two photos in the book “Lieutenant-Colonel John Haughton, Commander of the 36th Sikhs” that show Fort Lockhart viewed from Saragahri. One was taken outside the Saragahri post, and the other was taken from inside.

    Determining the N/S location where the photo taken outside Saragarhi post is something we should be able to determine to exquisite precision by looking at the alignment between a nearby object still available in the modern day photos, but is much closer.

    The bend in the Samana road close to Saragarhi is an excellent choice, since it skirts the north edge of the ridge, a feature which we can hope will not have shifted much in the last 120 years.

    The photo shows how sensitive the apparent left-right position of that road curve with respect to horizon features (such as Fort Lockhart and the edge of the cliff south of Fort Lockhart) is to the North-South location of the photographer. (The colored feature lines were obtained by superimposing “ground view” images from Google Earth of the road and horizon from various locations.

    The reason for this is perhaps clearer from the left image which shows how a line from the photographer through the bend of the road (blue lines) would shift as the photographer moves N/S along the ridge (magenta line).

    My initial rough estimate is that the position the photograph was taken from is close to the left (North) end of the magenta line, as the cliff lip east of Fort Lockhart appears to the left of the nearby bend in the Samana road in the 1897 photograph, but I believe I can refine this further with a few more iterations.

    The analysis showing locations visible to both forts places Saragarhi quite well in the East-West direction for a given position North or South, but is a bit fuzzy in the N/S direction.

    The period photos taken of Fort Lockhart from Saragarhi, however, can be used to accurately place Saragarhi in the N/S direction, but will be weak in the East-West direction.

    By combining the two sources of information, we should be able to confirm the position of Saragarhi post without the need to lean on the statement that it was at the “highest” point on the ridge.

    There are two photos in the book “Lieutenant-Colonel John Haughton, Commander of the 36th Sikhs” that show Fort Lockhart viewed from Saragahri. One was taken outside the Saragahri post,and the other was taken from inside.

    I’ve been using the first one to narrow down the location of the post, but wanted to check how close the location it was taken from was to the post.

    The answer seems to be – very close in the N/S direction.

    With some more cogitation, I may be able to answer the question ” how close” …


    I scaled the 2nd shot to match the first, rotated it -2.4274 degrees to line up the skyline (it seems that one or the other photo was not taken with the camera perfectly horizontal), cropped out the part that showed the “interesting” part of the horizon, and overlaid it with the first.

    I then made an animated gif (third attachment) to flicker back and forth between the two for a “blink comparison”[1] of the two.

    Besides the good horizon match, notice how well the two dark spots on the hill just below Fort Lockhart correspond in the two photos. If the two photos were taken from appreciably different N/S locations, then those spots would move left or right with respect to Fort Lockhart between the two photos.

    Alas, since those two spots are not likely to be in modern photos, we can’t use this to check against modern photos, but it does provide confidence that if we can line up modern images with the photo taken outside Saragahri post, we will not be far wrong in locating in the N/S direction, where the post was.



    [1] For two reasons:
    (a) the high point of the ridge is quite flat in the N/S direction – a 66 meter change in position
    for a eight ft change in height, per Google Earth’s estimate.
    (b) The absolute height is fairly uncertain in current global DEMs (though I cling to the hope that
    _relative_ height, which is all we need for the highest point, is relatively accurate for relative smooth, relatively flat terrain.
    Of course, as stated in (a), relatively flat terrain makes locating the local high point, touchy.




  2. Is Saragarhi in the right place?

    Leave a Comment

    Through our Saragarhi Live videos we’ve connected with many people from around the world who are interested in the battle and our forthcoming film. A continuing conversation with one of these has been fascinating and forms the basis of this post – pinpointing exactly where Saragarhi is and why this location was the best for it.

    In researching for the book “Saragarhi: The Forgotten Battle” we found several primary sources which indicate and describe exactly where the signalling post was situated, including the original document for the placing of the Samana posts after the second Miranzai expedition.

    The official history from the India Army Intelligence Branch, 1908, puts it simply that: “Saraghari was … situated on the the highest point of the range between Fort Lockhart and Gulistan.”

    We took this research to pinpoint the site of the post on Google Maps, right next to the Samana Road.

    But why was this the best site? Could Saragarhi have been located somewhere else?

    Richard Fowell, a heliography expert from California, got in touch with some interesting insight, which we are publishing here with his kind permission.

    Using digital elevation maps (DEM), Richard found not only where Saragahri was located (33.554º N 70.888º E), but all locations where a relay post in between could have been sited.

    Below are his calculations:

    “Various accounts place Saragarhi at various distances from Fort Lockhart and Fort Gulistan, and variously north or south of the road (or line) between the forts. Since Saragarhi has been obliterated, we can’t directly locate it with Google Earth. However, we should be able to locate Saragarhi definitively from topography. Based on the below, I place it at 33.554º N 70.888º E.

    I began my analysis at Michael Kosowsky’s free site His site uses Google’s computerized contour maps (digital elevation models (DEMs) to find the highest point within a given radius of a location, to show areas visible from a given location and height, and to find the bearing and range to other points.

    Fort Lockhart and Fort Gulistan are marked on Google Maps – I began at a point midway between the two. The “Contours” option at displays the ridge between the forts. As the 1908 official history places Saragarhi at the high point of that ridge,  I picked a point in the centre of the ridge and asked for the highest location within one mile. chose 33.554167º N 70.8875º E 6467 ft elevation. The areas visible from eye level at that point (6 feet above ground level (AGL) : 6473 ft) are tinted red in the first screenshot from the analysis here: Saragarhi Post.

    This puts Saragarhi 1.8 +/- 0.1 miles (as the heliograph flash travels) from both forts It is south of, and 60 ft above, the crest of Samana Road.

    Some published accounts place Saragarhi post north of the road – how do we convince skeptics?

    DEMs aren’t perfect. The DEM for this region was probably based on Shuttle radar data, likely 90 m horizontal spacing and 16 m (53 ft) vertical accuracy (90% confidence), with any points in between interpolated (an “educated guess”) from that data.  The Google Earth DEM for this ridge puts  the location above 22 ft lower (6445 ft), and says the ridge high point is 66 meters SW, and eight feet higher, at 33.553673º N 70.887908º E 6453 ft.

    We know Saraghari was visible from both Fort Lockhart and Fort Gulistan – we can use this as a check.

    The locations of Fort Lockhart and Gulistan are marked on Google Maps, with the buildings visible in “satellite” view.  I used to pick the highest point in each fort. The 1908 official history says the forts had 14 ft walls, so I set the height to eye level above the walls (20 feet AGL). provided these views:

    Fort Lockhart +20 ft AGL: latitude 33.556353º N longitude 70.919012º E elevation 6592 ft above sea level

    Fort Gulistan +20 ft AGL: latitude 33.557208º N longitude 70.856536º E elevation 6086 ft above sea level

    I overlaid these views, with the top view set to 50% transparency, to get the second screenshot, where locations visible from both Lockhart and Gulistan are darker red.

    One interesting observation from the intervisibility map is that the heliograph station did not have to be placed between the two forts to communicate to both. [1]

    I sketched in blue lines to show all the locations between Lockhart and Gulistan that were visible from both. 
    I’ve marked an attractive alternate signaling site north of the road, about 1 km NW of Saragarhi post. Militarily, this northern location has the disadvantage that snipers at Saragarhi would (slightly) overlook it, but Fort Gulistan suffers the same disadvantage with respect to the slopes west of it.

    The northern point is 25 m lower than Saragarhi, so it is clearly not the highest point on a ridge, but it is a high point on the ridge. Are we putting too much weight on the adjective “highest” in the 1908 account?

    We can show that our location is correct by combining information from the DEMs with photogrammetry from photos of Saragarhi taken shortly after the battle. That will be the subject of my next communication.” overlay images are Copyright 2017 Michael Kosowsky. All rights reserved. Used with permission.

    [1] A contrarian choice would have been a signal post at Samana Suk, west of both forts. It would be within 5 miles of both forts – point-blank range for a heliograph. It would also offer advance warning for any incursion from the west.














  3. How did Saragarhi signaller Gurmukh Singh configure his heliograph?

    Leave a Comment

    Guest blog by heliography expert Richard Fowell, California

    During the battle of Saragarhi, Sepoy Gurmukh Singh would have signaled Fort Lockhart with his heliograph configured with a single mirror early in the day, and with two mirrors later in the day.  The first attachment shows an 1886 illustration of the two configurations, and below that is a modern photo:

    The time of day at which Gurmukh Singh would have added the second mirror would have been 1:50 PM local solar time  which would have meant his noon signal would have been with the single mirror, and his 3pm  and final messages with two mirrors.

    Here is the reason why:

    When the sun and the target station are on the same side of the signaler, a single heliograph mirror will suffice. However, as the angle between the target and the sun increases, a single mirror catches less and less sun, and becomes ineffective. Hence, from the earliest days of the heliograph, a second mirror was provided, and the signaler was taught to use it when the angle between the target and the sun was too great.

    The 1886 and 1889 signaling manuals did not provide specific guidelines about what angle was “too great”, but the 1889 “Catechism on the Manual of Instruction in Army Signalling, &c by Edye, L., Rhodes, E” states the use of the second (“duplex”) mirror when the angle between the sun and target station seen at the mirror was greater than 120 degrees.

    Later manuals said 90 degrees rather than 120 degrees, but in 1897, it seems most likely that Gurmurkh Singh would have been taught that 120 degrees was the angle to switch configurations.

    Given the date of the battle of Saragarhi, and the positions of Saragarhi and Fort Lockhart, we can compute that angle as a function of the local solar time of day – the details and websites used are below:
















    Note: I don’t know what time system Colonel Haughton and Major Des Voeux were using when they reported various events of the battle. If they were using a standard time, rather than a solar time, their numbers would be slightly different than those in the spreadsheet, but in any event, they should
    not differ by more than an hour, which means that the switch point would be somewhere between 12:50 and 2:50, meaning that the noon signal would be single-mirror, and the 3PM and final signals dual-mirror (duplex).

    The difference between the British military time and Saragarhi solar time, if any, may have been much less than 1 hour.
    In 2017, the neighboring standard times differ from Saragarhi solar time by 0.226 – 0.772 hours ( 14 to 46 minutes), as follows:

    Greenwich Time + 4.500 hours: Afghanistan Standard Time (2017)
    Greenwich Time + 4.726 hours: Saragarhi Solar Time (always)
    Greenwich Time + 5.000 hours: Pakistan Standard Time (2017)
    Greenwich Time + 5.500 hours: India Standard Time (2017)


    Richard A. Fowell