Category Archives: Juggernauting
I said there’d be no more Juggernauting posts until the video(s) were released, but I’m a filthy liar.
I finished the rough cut of the five-reel premier version reconstruction a few days ago, and not to toot my own horn, but I’m very pleased with it. It clocks in at 58 minutes and 33 seconds, not including credits or whatever introductory text might be prepended, but it may gain or lose a couple minutes in tweaking the animation speed and title length. It will take several times going over until the flow is right. 30 minutes and 40 seconds are actual moving pictures, the remainder is a photo reconstruction.
Of the lost reels, reel one was the easiest to recreate. I had seven separate stills to work with from it, not to mention a lot of easily repurposed footage from later in the film. I only had to resort to footage from other films for the poker game scene. It’s such a major part of the story that it has to be represented, but I’ve no stills of it nor any material elsewhere in the film that could be made to resemble it even through the most creative use of cropping and matting. I took a shot from the poker game in The Girl in the Arm-Chair and composited William R. Dunn’s head onto Mace Greenleaf’s body. I would have much preferred to use only Juggernaut material, but failing that, it works nicely. (I should clarify, I don’t have any stills of the game itself. I do have two stills of the fight that follows when Philip discovers he’s being cheated.) The long happy ending’s hospital scenes in reel five also requires some outside footage. Just like in the 2012 reconstruction, it comes from Wanted:- A Nurse, with Anita Stewart and Earle Williams filling in for Sidney Drew and Ethel Lee.
Biggest trouble is reel four. For reel four, I only have two production stills, and they’re actually from the same part of the same scene from different angles. (More accurately, only one of them is a production still, the other is a frame enlargement. The still photographer stood about six or eight feet to the right of the cameraman, evidently.) That’s actually the main reason I sat on the footage so long — I was hoping another reel four still or two would turn up, and I didn’t want to go ahead with what I had until it became clear nothing else was going to surface. As I said before, that happens really quite often. The original Everette True video was scrapped at about 90% completion when I found a better quality print. More recently, The Sawdust Ring had been out about week when I acquired a print that’s largely identical and of inferior quality to the one I used, but it does have some footage that my Argentine print doesn’t. It does help, though, that the surviving last reel of Juggernaut is actually from the four reel abridgment and as such it contains a bit of footage from both the original fifth reel and the fourth — about 80-20 — so while I struggled with scrounging up images to use, reel four is actually the longest of the reconstructed reels thanks to the three minutes of surviving footage towards the end.
The five-reel premier version will be tinted. I know it would have been tinted originally, but I’ve found no record describing just how it was. The tinting scheme I used is based on the two Vitagraph films I have on nitrate, Her Faith in the Flag and a brief fragment of an unidentified circa-1915 Maurice Costello drama. It’s a very ordinary scheme, nothing unusual in what color is used for what situation. The only thing of note is the intertitles, which are tinted purple. The main title, being a Blue Ribbon release, should obviously be blue. The amalgamated or “watchable” version I’ve spoken of before, with the short happy ending and all the abridgments left as-is, will be black and white.
Speaking of the titles, while the watchable version will retain the original titles (excepting those in the reconstructed reels, of course), the premier version’s will all be replacements, for this reason: The picture quality takes a hit with each generation removed from the camera negative. It will get grainier, which can be corrected to a degree with judicious temporal and/or spacial smoothing. It will get fuzzier, which sharpening filters of various sorts can address. It will get more flickery, which — provided your image has got enough depth to start with — can be nearly eliminated in all but extreme cases. But the most difficult thing to undo is cropping. For the picture, you can build up something of a background if the camera tilts or pans, but that’s often unsatisfactory and very little else is possible. Titles, though, those can be un-cropped simply by redoing them. Example, we want to take this title:
…which is closer to how it would have originally appeared. And there’s no trick or magic to that — it’s just loading the original image in Photoshop and having the patience to paint in what’s missing. All the premier version’s titles have been brought back to something close to their correct framing.
The reconstruction dialogue titles are all exact reproductions of the original text with one exception, that being Louise’s first line, which I changed from “Why are the papers so antagonistic?” to “Father, why are the papers so antagonistic?” to make it more obvious who she is. It’s just after the time jump and we haven’t been introduced to Louise yet or filled-in on what Philip has done since Viola’s death. Without making it clear, it’s too easy to mistake her for his second wife. The descriptive titles are less exacting but hew as close to the synopsis text in the exhibitor’s handbook as possible.
It will take a while to score The Juggernaut. For one, it’s a feature-length film and I’m rather out of practice with those. With a short, if you like a piece and it fits the scene, you use it, because it’s probably the only scene of that type in the film. With a feature, there are a lot of similar scenes that that piece might fit, and you want to make sure it’s used to best effect because you can only use it once. Unless you’re doing character themes, of course. Like, in The Soilers, all of Clarence’s scenes are scored with Milady Dainty, but each is snipped from a different part of the tune — you can’t just endlessly repeat it. Themes are a good way to handle a film without a lot of action or films where the action is one-note (like The Soilers). Two, in photo reconstructions, there’s implied motion but no actual movement for the music to follow. It falls on the music itself to suggest action.
Finally, for the next video, I went with my first inclination and threaded-up Doll-House Mystery in the scanner. I was also considering Broken China, since I really wasn’t aware until recently how rare my print of the film is. The one-reel abridgment is reasonably common, but mine is the original two-reel version, which is considerably less so. But I’m just not in the mood to work on a film I despise. Dancer is all right but I like Doll-House quite a bit. I usually process the film as it’s being scanned — saves time that way — but I’ve been entirely focused on the The Juggernaut and haven’t done anything but let 604 gigabytes of raw image data build up on the capture drives. I’m not looking forward to tackling that.
In fact, let’s not. Let’s start scoring Juggernaut. Doll-House can wait.
I can say with certainty that the short happy ending of The Juggernaut was filmed after-the-fact for the 1916 British release, which was handled through Gaumont. Like the later American re-releases, it was shortened from five reels down to four. Most of this I’m sure came from ditching the hospital scenes, which from the description must have been several minutes long, but you can see evidence of shortening in many places.
In the existing cut, in the scene where Philip learns that Louise is on the doomed train, Philip tells his secretary Reynolds to “get busy on the phone and I’ll try to head them off,” then dashes out of the office. Reynolds picks up the phone and calls Brandon, but is told “It’s too late.” In the original version, Reynolds speaks first, crying “Stop the express at all costs!” before hearing Brandon’s reply.
That’s repeated several times in the last reel — cutting everything but the conclusion from dialogue scenes. It’s actually really well done, and unless you’re reading along with the script (as I’m essentially doing), you don’t notice anything is missing. If you’re an editor tasked with abbreviating a film by about a reel, that’s a great way of accomplishing it without altering the overall narrative.
The last reel is assuredly from the four reel abridgment. The second reel is a different story. The surviving footage follows along exactly with the synopsis, except for one scene that isn’t edited so much as it’s missing entirely. On the screen, we only learn about Philip’s father’s death after the fact when Viola’s mother sees the newspaper headline. That happens in the text, too, but first Philip is supposed to find his father’s body. I know the scene was shot because I have a production still from it. The second reel varies in condition, with some parts that look rather good and others that are considerably decomposed. I suspect the preprint was verging on unsalvagable at that point and they just snipped the whole scene out.
All of the film is transferred and the video has been processed and finalized. It actually has been for the better part of six months. I was really just sitting on it and waiting to see if anything more — in terms of stills or information or whatever — should turn up, as it so often does when you near the end of a project. It hasn’t though, so I’m moving into the next stage: reconstructing the missing footage.
On the 2012 DVD, which I’m sure all of you have, you’ll recall that I used all the existing last reel/short happy ending footage in the reconstruction. The other two endings were included in a separate bonus short. I’m not decided, but for the new edition, I’m kind of leaning toward doing two reconstructions: one would be of the New York premier version, the other you might call the “watchable” version. The premier version would be my best stab at re-creating The Juggernaut as it was originally released, but it probably won’t be for the casual viewer. It’s jarring to switch from live action to panning and zooming on stills. Reel two isn’t so much a problem, since again, there’s at most only a minute missing. Reel five, with its frequent abridged scenes, is going to be very choppy indeed. It’s of interest to me and possibly a handful of others, but for those that just want to watch a movie, it’s probably best to leave all the abridgments as they are and interrupt the action as little as possible, even if it doesn’t accurately reflect any particular incarnation of the film (being a melding of both the five and four reel versions). That would only leave the long happy ending unrepresented, but it could be done in a bonus video.
I imagine this will be the last Juggernating post until I actually get around to reviewing the film, but who knows? That will still be some time yet. In other news, a hint as to the video coming out next week or so: Bessie Love joins the circus.
Recently went on a little trip that took me through New Jersey. Decided to take exit nine off the turnpike, get out, and have a little walk around. Why don’t we head across the South River, into the town of Sayreville? There’s a little nothing of a park off South Minnisink Avenue— not much beyond a small field and a few trees — but if we walk all the way to the end of it, what do we find?
A fence? Nuts to that.
Well now, here’s a familiar looking pond. Seems that I recall something significant happened in this pond about a hundred years ago. That sandy beach, too — why, even now I can see William R. Dunn running across it, kicking up the sand as he skids to a stop. Let’s hike up around the north edge.
What’s this? Why, those tracks look like they might be from the old Raritan River Railroad. If I squint, I can just make out an antiquated train rolling over them, with Anita Stewart in the window, looking pensive. What might these tracks lead us to?
And there’s what we came to see. That’s what remains of the climax of The Juggernaut (1915) — the weathered, rotten stumps of the trestle that failed to carry the train across the water.
In other news, work is well underway on the new reconstruction of The Juggernaut, but it’s a big project that won’t be finished anytime soon. The extras will be done first: A Railroad Smashup (1904),
A Mother’s Devotion (1912) The Firing of the Patchwork Quilt (1911), The Black Diamond Express (1896), plus one or two other short train actualities. Expect to see them in the coming months.
Before they began producing their own films, Vitagraph exhibited Edison pictures. The very first they presented publicly was The Black Diamond Express (1896). It’s a simple actuality — nothing more than a train approaching the camera — but it was wildly popular. Difficult to imagine today, but then, trains were the modern marvel. Also difficult to imagine are smashups — a kind of carnival sideshow common at the turn of the 20th century, where antiquated railroad engines were purposefully wrecked for the audience’s entertainment.
Filming a smashup wasn’t a novel idea — witness Edison’s The Railroad Smashup (1904) — but incorporating one into a feature-length narrative film was.
Vitagraph had purchased four decommissioned engines from the Long Island Railroad and several old passenger cars. Each train was destined to be destroyed in some manner. The first two met in a head-on collision, in standard smashup fashion, in The Wreck (1914). The fourth speeds into a forest fire in The Ninety and Nine (1916). The third, retired Long Island Railroad Engine #56, would find itself plunged into the icy depths at the climax of The Juggernaut (1915).
The train wreck was filmed on September 27th, 1914. On the northern edge of the Hercules chemical plant in Sayreville, New Jersey is a large man-made lagoon locally called Duck Nest Pond. Running hard beside it are Raritan River Railroad tracks. Vitagraph constructed a short spur line off these, which made a wide turn onto a trestle that spanned half-way across the pond.
In the story, the railroad has been dangerously neglected. The bridge, we find out only too late, is unsafe to cross. At first, the wobbly piers support the train’s weight, but then — crash!
The passenger cars were filled with dummies. The only live human aboard was the engineer, who jumped off some distance away after wiring the throttle down. The train was meant to cross the trestle at around 15 miles per hour, but between the engineer’s departure and the train reaching the spur, the wire had worked itself loose. Unlike the ersatz dereliction of the bridge, the engine was legitimately in disrepair and the throttle dropped into full open. I’ve heard it said the train topped-out at anywhere from 30 to 60 miles an hour before it reached the trestle. From the footage, I might believe 30. It’s a bit over-cranked, which makes it seem to move slower than it is, but there’s simply no way that train is moving anywhere near 60mph.
The piers were rigged with explosives to ensure that the bridge failed on cue, but the train was moving so fast that, even as the center of the bridge was collapsing, it still managed to cross it and go sailing off the end — narrowly missing one of the camera towers. It was expected that the boiler would explode when it hit the cold water, but it didn’t. Pyrotechnician Herman Rogers assisted it with a couple sticks of dynamite.
Once the dust settled, the dummies were replaced by living extras, to be filmed struggling out of the wreckage and swimming ashore. Easier said than done. There were lifeguards stationed, but it wasn’t always easy to differentiate those who were pretending to drown from those who actually were drowning. There were no major injuries, but several people had to be rescued.
Most attributed their cramps to the chill of the water, but the temperature is the least of Duck Nest Pond’s problems. The Hercules plant was a major producer of nitrocellulose, among other chemicals. Indeed, the film The Juggernaut was shot on may have come from Hercules. But they kept their waste on-site in unlined landfills, which leached into the groundwater. The area surrounding the plant is highly contaminated with tertiary-butyl alcohol, carbon tetrachloride, and chloroform. The place was and is toxic. Today, it’s an EPA Superfund site.
A century later and much of the scene is still there down in Sayreville. Vitagraph just abandoned everything when filming wrapped. The spur is gone, but the main tracks are still there. The pond is still there. Rows of stumps soldier off into the water — the remnants of the piers that made up the trestle — they’re still there. Even the wreck itself was left to rust away for decades — it wasn’t until 1938 that it was pulled out for scrap iron.
Unrelated to this post or The Juggernaut, a new video will be going up soon. Not going to say what it is, but it’s a personal favorite that I’ve written about before. Maybe you can guess.
Continuing the disorganized musings on The Juggernaut (1915) I began in Part 1:
There are three different descriptions of the film’s ending: there’s the downer ending, in which Louise dies; there’s the long happy ending, in which Louise slips into a coma and is hospitalized but eventually recovers; and there’s the short happy ending, in which Louise briefly loses consciousness but wakes up moments after being dragged ashore.
It’s almost certain that the New York premiere had the downer ending. The synopses provided by the producers to trade magazines didn’t always correspond in every detail to what was actually released, but the downer ending as described is well attested by contemporary reviews. The other two endings — those are more questionable.
The long happy ending is the one the novelization in Motion Picture Magazine uses. Now, as was the case for all of their novelizations, this was written well before the film was actually completed (perhaps before filming even began) so that it could be released concurrently. It was more or less loosely based on the script, rearranging things and adding more dialogue to make it a better read. It could be that the long happy ending never existed outside of this novelization — except for the fact that it’s alluded to in some reviews of the film. It’s my assumption that the long happy ending did exist and that both it and the downer ending played at the same time in different markets (urban/rural? domestic/international?).
Of the three, the short happy ending is the least attested. Indeed, I haven’t actually seen any reviews that reference it at all. I’d be more tempted to doubt its existence, if it weren’t for the fact that it’s the only one that survives.
Every print that exists today of what’s ostensibly the fifth and final reel of The Juggernaut originates from the same nitrate source — that’s quite clear from the identical decomp visible in the pre-print. I say “ostensibly” because, while I don’t deny that it’s the last reel, I don’t think it’s the fifth reel. I think the footage that survives comes from either the 1917 or the 1920 re-release versions, which were shortened to four reels and may have been edited in other ways as well… say, a new ending. It does show some evidence of condensing — at the start, we see the tail end of the trial scene, which should have been contained entirely in reel four. There are also a number of shots that seem to be abbreviated. The shot when then car pulls up to the dock, for example, is something like 15 frames long and cuts just as the car door begins to open. Without slowing it down, it’s impossible to even tell what’s going on. While I imagine this was always a quick shot, I suspect it originally showed more of the car’s approach and the two men actually getting out of it, which would establish continuity for the next shot.
The reviews of the long happy ending are critical of it needlessly prolonging the film beyond the climax instead of going out while the audience was still hot. It’s my theory that, if this footage is from a re-release, then it shows that the producers heeded the critics’ opinion and the short happy ending was created as a response. And it would explain the general absence of the short happy ending from reviews. When they existed at all, reviews were quite cursory for re-released films that had already been reviewed at length during their original run.
Assuming I’m right, then the question is was the short happy ending assembled from footage that had already been shot or was it filmed later? I can’t say, other than, given the editing, it’s quite possible that some time had elapsed between the shot of John pulling Louise out of the boat and the shot of him laying her on the ground. The background extras don’t appear to carry over between those two shots. The only featured players in the final scene are Earle Williams and Anita Stewart. Williams was still with Vitagraph in 1920. Stewart was not, but she was in 1917. Was the short happy ending filmed in 1917 for the that year’s re-release?
There are a handful of films that, for whatever reason, I’m intensely attracted to. I spend years trying to track down as much material relating to them as I possibly can. The Juggernaut (Vitagraph, 1915) is one of those films. This post is the first of what will probably be a short and sporadically updated series I’ve decided to call “Juggernauting” detailing the present state of my obsession.
The Juggernaut — “a story of modern life” about greed, corruption, betrayal, and retribution (and a full-scale train wreck!) — was one of the biggest films of 1915. In terms of box office receipts, only The Birth of a Nation was more successful. It played to sold-out houses for weeks on end in its first run, and was re-released twice, in 1917 and 1920.
Today, The Juggernaut only partially survives. Some time ago, I made a reconstruction using the intact fifth part, a fragment of the fourth part, and several production stills from the other three parts. I was rather pleased with it, and to date, the DVD has been one of our top-ten sellers. It’s outdated now. I’ve got many more stills, I’ve got an exhibitor’s handbook that details every scene and quotes many of the titles (previously, I knew the story in outline, but had to make “educated guesses” about the specifics), and the most exciting development of all, I’ve obtained a complete copy of the second part.
The 2012 reconstruction ran for almost 30 minutes exactly. I haven’t begun another yet — I’m still in what I call the “amassing stage” — but I now have 31 minutes of actual footage from the film alone, never mind all the stills I have to use in recreating missing scenes. For the next reconstruction (let’s tentatively call it the 2016 version), I anticipate a new running time of around 50 minutes, which is not far off the film’s original length of 65 minutes.