Location 1: Havice Valley, Mifflin county, about 40.79714,-77.46911
With a perfect day weather-wise, I decided to borrow a friend’s metal detector and investigate a couple suspicious spots along Havice Valley Road in Price Gap. One dated from about a year ago, another from just last week. Both were close by Havice Spring, shown on most topo maps. On the east side of Havice Valley Rd in Price Gap, there is a ledge that looks a lot like RoW, and which curves to the south and east around the corner of the mountain.
On the west side of HVR, there is another ledge, built of substantial rocks at the bottom near the stream, which climbs steadily up the mountain to the south and west of the gap.
Despite the promising appearances, the metal detector seemed to deny there was railroading involved. I found metallic junk, but no “good iron”. I regretfully have to place these two leads on the Unlikely list.
Location 2: Flat Hollow between Pitch Pine Ridge and White Mountain , Mifflin county, about 40.79458,-77.44542
Refusing to be discouraged, I continued east on HVR to the junction with Little Poe and Strong Mountain roads and took SMR to the south. This is a fairly rough road (a “drivable trail” in DCNR parlance) and requires good ground clearance, if not 4wd. It provides an impressive view of Price Kettle and taste of the north end of Kish Valley.
I reached the flat just short of Flat Hollow Rd and parked near the vernal ponds on either side of the road. Based on Kline’s info and our earlier survey, it seemed Duncan’s railroad should have reached this point. I first crashed around in Mtn Laurel near the junction with FHR and found nothing except Mtn Laurel. I then followed my original plan, heading ENE on the trail that crosses SMR at the ponds (40.79509,-77.44543).
My original goal was the point about 1/2 mile E where the trail diverges 90 degrees. It seemed logical to look beyond the bend. However about 1/4 mile out I noticed fairly clear rocky areas to my right and decided to have a lookabout. I didn’t immediately notice any good grade, but I did see a very large pine tree. Following my theory about large pine trees, I examined its base and poked about with the metal detector to see if it was on the RoW. It wasn’t. A short distance away, however, the detector sounded a fairly loud alarm. I quickly dug up a broken four-bolt joint bar. Hum. I don’t think this is a natural formation!
I soon turned up a track bolt which allowed me to get the bearing of the RoW, which I surveyed a distance to the east before the brush became too thick.
So this raises an interesting question:
Why a second joint bar on what Kline says was a wood-railed gravity tramroad?
Returning to the big pine tree, I surveyed a distance to the west until I reached a pine thicket which exceeded my determination. I suspended the track and cut back over to open trail. Naturally I was right on a branch of it that I had previously explored; that always seems to happen!
On the way back to SMR I tried to check a spot or two on the RoW, but the swampiness and thickets atop the flat are pretty discouraging. I decided to try the west side of SMR.
West of Strong Mountain Road
I walked west on Flat Hollow Rd, trying to avoid Mtn Laurel. I first checked on the south side of the road due to something I’d seen on aerial views and found nothing.
I then tried the north side of the road where it seemed RoW should be. In a fairly rocky area, I found a very large pine tree. Examining its base with the detector, I found exactly nothing. Not far away, the detector alerted me to metal (are you having deja vu yet?) and I pulled out an encrusted spike. Squinting, I figured I sort of had the bearing of the RoW and began to record a track.
After a while the track entered a thicket and swampy area and became vague. I expected to find some timbers in the dampness, but no luck. Crossing the associated small stream, I couldn’t pick up the RoW again. Using the detector, I located a signal and pulled out an interesting fragment of a chain link. Knocking off the thick corrosion, its fibrous nature is very obvious… the link was clearly hammered from wrought iron by some long deceased blacksmith. I decided that was enough excitement for one day and began a long slog back to the truck through very thick Mtn Laurel.
Exeunt
I decided to exit via Flat Hollow Rd, which was a good choice. Approaching McNitt Gap, the valley narrows considerably, limiting where the RoW could be. I stopped for a quick check (about 40.78785,-77.46058) and saw probable RoW plunging to the stream just alongside the road. The detector showed something substantial there, but my limited energy didn’t allow me to determine exactly what it was. Suffice to say it seems very likely that Duncan came all the way to McNitt Gap, if not actually into the gap, which agrees with Kline’s map from Book #2 page 217.
Location: Mifflin county near Poe Mills on Penn’s Creek (approximately: 40.82255,-77.40535)
Ben Kline’s books say that John Duncan logged both Swift Run and Rocky Run in the very early days of steam powered logging railroads, perhaps 1889 to 1899, well before he began his extensive operations at White Deer, PA. While traces of right of way do seem evident along Swift Run Road (aka Paddy Mountain Rd, aka Havice Valley Road), Rocky Run has always seemed impossibly steep and, er, rocky for a logging railroad.
Last week a group of us were out exploring in the general area and got to puzzling over Rocky Run on our way back out of the woods. Was it really ever accessed, or was Kline wrong? First, we examined the south side of SRR to the west of Rocky Run. Was there any way a route might have switchbacked off of Swift Run and wrapped around the end of Pitchpine Ridge to get higher up Rocky Run? The short answer: No.
We then decided to take another look at Rocky Run itself. We stopped at Point Lookout Camp and piled out. We gazed upon the small stream heading steeply uphill. We gazed at the big and jumbled rocks. We gazed at the slope. We muttered discouraged words and got back in the car!
Heading N/NE back towards Poe Mills, we studied the surrounding terrain, pondering how else Duncan’s men might have bypassed the lower portion of Rocky Run. We shortly realized that there are other gaps leading off Swift Run in the general direction of White Mountain and Rocky Run. One dead ends into White Mountain Kettle and thus seems a poor bet; a kettle is usually pretty snugly contained. The next gap to the south contains White Mountain Ridge Trail, which sounds much too developed. Probably a donut shop up there. But the gap to the SSW… hmmm.
We therefore parked along the road and attempted to cross Swift Run, made difficult by the fact that it’s fairly swift, not to mention wet and cold and wide. Spotting a small collection of hardware on display along the stream inspired us however, and we found a crossing and began exploring up the gap.
Only a few steps beyond the camp, I spotted some curious terrain that looked like it just could be grade. I decided it was worth proceeding, as the others fanned out in various directions. A short distance upstream I glanced up at a lengthy and well weathered timber running along my path and let out a whoop–two spikes were sticking out of it!
I gathered several others and we explored upstream. It seemed pretty likely there was track some distance, but due to the late hour and failing light, we decided to return another time.
The Second Visit, Leading to Other Surprises
On January 2 the weather was blustery but reasonable, so I called upon a team member to see if he was up to knocking out a survey of the new route. “Bring the metal detector!”, I urged.
Arriving at the unnamed stream, we started sweeping with the metal detector and quickly started pulling iron from the ground. Our first find was an interesting piece that looks like a hook off the end of a chain, perhaps part of a log car or even a horse’s trace chain. Typical early spikes (long and thin) were everywhere. We proceeded upgrade, confirming the route by finding traces of iron all over the place, including several places in the long timber mentioned above.
Eventually we reached a rather forbidding wall of earth where we had stopped on the earlier trip. It seemed impossible that they could have crossed it. Doing an “oh what the heck…” we climbed beyond it, only to find iron again immediately beyond. Turns out an earth slide in the intervening 120 years thoroughly buried the route, but it continues quite nicely! This is shown as ‘Earth Slide’ on the map.
Finally a rock pile did bring us to an abrupt halt. Above it was a strip of iron several feet long, with a chamfered end and regular screw holes; possibly a runner off a logging sled? It didn’t seem heavy enough or hammered enough to be strap iron for rail.
We continued a short distance uphill, theorizing that this region was perhaps used to skid logs to cars waiting just below. A nice signal from the detector set us to digging only to find… a horseshoe, complete with nails. Hmmm… Perhaps this region was used to skid logs to cars waiting just below!
Concluding that anything beyond really falls to HistoricHorsePoweredLogging.com, we headed back downhill, starting a GPS track at the upper reaches of what could have been track. The total survey was a hair under 1/2 mile, descending from 1515 feet at the horseshoe to about 1060 at the camp, giving an average 17% grade! Seriously.
But it Gets Better!
Returning to the car, we fortified ourselves with homemade apple pie and headed for Point Lookout Camp. One of those moments of: I know it’s impossible, I just want to see if it’s possible. With the GPS loaded with a track visible on aerial views, we started up a trail behind the camp. We know they didn’t go exactly this way, we gasped as we ascended the steep slope. After a bit it leveled off and we skirted the west face of a bulge of White Mountain.
Eventually it seemed we were coming into a leg of a hollow branching off Rocky Run. With some bare rock to the south, it seemed like a good place to do a transit of the hollow. Shortly after crossing a dry streambed, I glanced up and gave a start. What is this? My companion quickly came up and joined me in exploring a rock feature that was definitely constructed intentionally. We soon determined it couldn’t really be right of way. But it sure could be a large rock slide, in fact a funnel where several slides joined, thus I proclaimed it: Funnel.
Crashing about in the brush, we tried to determine where the logs went next after they were Funneled. It didn’t take long to realize a substantial trench extended north west. Perhaps they skidded the logs along here using horse power. The trench extended. And extended. After about 250 yards, we entered an area where rocks tumbled steeply down away from us to Rocky Run. The trench stayed to a gentle slope along the hillside. A good buzz from the detector set us digging only to turn up… a horseshoe! One could almost believe they skidded logs along here using horse power.
Finally the trench took a steeper dive down the hillside, and the metal detective hollered that we seemed to have a loading point. Perhaps if we looked below it we would find… iron? Of course, piles of it! But here we are, on the upper reaches of Rocky Run, where it is plainly impossible to be, yet we have iron. *sigh*
We made the decision to head upstream for a fixed time due to fading light, then survey downstream with the GPS, attempting to record the route out of The Impossible Rocky Run. Turning upstream, we looked in disbelief at where we were going. Is this possible?
Scouting ahead with the metal detector showed iron. We continued upstream, marveling at the rugged terrain and the tumbling stream that they somehow surmounted. The iron continued, and eventually we started finding scattered timbers containing traces of spikes.
After a bit it seemed we were through the worst of the gorge and it was obvious the sun was at the horizon. We decided we better head downstream, so we activated the GPS and retraced our steps. [Later study shows we were approximately 1200 feet downstream of Bear Gap Trail when we stopped]
With the detector to affirm our recollections, we returned to our starting point and continued beyond. Shortly after, the ground began to plummet and it seemed unimaginable to continue!
Yet the route is there, clinging to the south the east walls of the gorge. At one spot, we spotted a crude but unmistakable fill constructed of large rocks.
Right along here we encountered a fragment of splice bar, which suggests they had T rail here.
There is so little shelf along here that we figured they must have been substantially “cribbing up” one rail above the hillside. Sure enough, we eventually spotted flat rocks stacked together which must have supported the outside rail. It would take a braver engineer than me to run much along this stretch of track!
Eventually we came into the northerly stretch of the stream and it became apparent that we were headed directly for Point Lookout Camp. What had seemed an impossible grade from below was just another day at the office for Mr Duncan’s gents. From this angle it was possible to see a ledge through the Mountain Laurel leading right out to the empty concrete pond in from of the camp. Examining beyond, we could even find the route on the other side of the present day road. Where the road obviously cuts through a bank just adjacent to Pt Lookout Camp, the grade hugs the hillside closer to Swift Run. Following it along the road, it is rather obvious (now) where the switch for this line diverged from the main near the road bridge over Swift Run. Doh!
This track also ended up about 1/2 mile, descending from 1468 to 1255. That gives an 8% average grade, surprisingly “modest” compared to the other hollow.
Conclusions… and Questions
Looking back at the Kline books, it seems obvious and logical that much of Duncan’s operations out here were wildcat, i.e. gravity runs of log cars down the hills. How far up these grades they went with the empty cars pulled by the locomotive (an early Climax class A) is debatable; I have a hunch they went to the mouth of these particular gaps and not much beyond, leaving the poor draft horses to lug the empties up these grades into the woods and skid logs to them.
Although most of the spikes we found were long and narrow, we did find a T-rail style splice bar and some short and stubby spikes that suggest iron T rail. Also, the timbers we found with spikes in them are curious: the spikes occur in groups of two. That’s a style we’ve typically seen where we think they had iron rail spiked atop longitudinal logs. A wildcat road from 1889 would seemingly be using entirely wood (4×4 or 6×6) rails, or maybe wood with strap iron on top, but we can’t see how the hardware we found fits with that.
We did learn one big lesson today: Don’t doubt the determination of these people, even in the earliest days of operation. And second: For best results, carry a metal detector!
With the help of the GPS, we returned to our previous stopping point and resumed surveying. The route was pretty clear although sometimes disrupted by a water crossing. We found good tie depressions in many areas. Although the ground got steeper and steeper, the builders did not give up for an impressively long time. It actually seems they surmounted one of the steepest areas, although approaching the saddle at the top of the valley it probably gets way steeper.
Near the end, the RoW encounters what appears to be a skid road at right angles. Shortly thereafter, the RoW seems to wind down and devolve into a couple of diverging log slides. One goes off to the east on relatively level ground, and one climbs steeply up the face of the ridge to the north. That seems to have been a loading point and the end of line.
We added 0.49 miles to our previous survey, and are happy to cross this branch off our list.
Here is a final track of all the Little Shingletown branch we managed to find:
If you’re interested in exploring history in remote areas of Pennsylvania, you should be concerned about the current push to legalize Sunday hunting in Pennsylvania. If passed, it would mean that people like us who like to explore are at substantial risk of death due to injuries inflicted by careless hunters. We feel that it is only fair that non-hunters have one day of the week that we can safely move about the woods. Many hunters I have personally spoken to agree, and believe the Sunday hunting is primarily aimed at drawing more out of state hunters.
The argument has largely been put in financial terms: the state believes they can make nearly a billion extra dollars by opening Sunday hunting. First of all, that seems highly unlikely. Second of all, fixing the state budget woes by endangering the few citizens who get off their fat asses and do something seems absurd and unfair.
The Pennsylvania Farm Bureau is OPPOSING the Sunday hunting bill. Please check out their website for details. You can also visit the Pennsylvania General Assembly Website to contact state legislators. PLEASE take a minute to send an email or call your representatives and VOICE YOUR OPPOSITION TO SUNDAY HUNTING.
Location: Livonia, Centre county (about 40.9712, -77.2987)
With leaves falling from the trees and snakes and bugs becoming less of a problem, it seemed like a good idea to spend a sunny Sunday afternoon near Livonia. I headed out Rt 192 and stopped at MacNeal Orchards to get some direction on where to find more right of way.
Based on their input, I headed south on Stover Gap Rd a short distance until a shale bank was visible on the west side. Parking there, I headed behind (west of) the shale bank, picking up some logging roads, and angled towards Elk Creek. Before long I was at the base of the hill approaching the stream. It seemed like there was sort of a grade, but it wasn’t very distinct. Continuing west along the base of the hill, I crossed a stream and suddenly I was on pretty distinct grade. I started a GPS track and headed west. Very soon I encountered ledges leading up the hillside to the south, but I decided to continue on the bottom track for a while and explore the ledges on the way back. The grade continued pretty clearly for about 3/4 of mile, then it got a bit vague. I left the GPS and explored ahead, but was a bit concerned I might not find it again. Plus I wanted time to head up those exciting ledges! I halted the track and headed back for my starting point. Based on discussions with the MacNeals, it seems this route continues at least as far west as Wohlford Run and the associated gap, so there is plenty of this level route left to survey on a future visit.
Back at my starting point, I examined my options. A very short stub starts up the hillside and stops after perhaps 150 feet. This appears to be either a false start of a route up the hill, or perhaps a loading ramp location. This is marked on the maps as waypoint “Switchcrap”.
I started climbing the next ledge up the hillside. It rose quickly, making me question how much I’d like to pilot a 20 ton locomotive down this hill after a rainy summer and fall like we’ve had here this year. At the top of the steep climb, it swung to the south with a nice rock fill… no question what this was built for. It then climbed into a bit of a hollow with clear tie depressions in places. Eventually the RoW intersected a dirt road which I am told is gated where it comes off Stover Gap Rd. I poked around a bit and felt the RoW probably continued nearby, but it wasn’t immediately obvious where. I suspended the GPS track and headed back with about 1/4 mile surveyed.
Back at Grand Central Station, I started a new GPS track on the uppermost ledge. This one was disconnected from the others, which seemed odd as it seems the most likely route down from ‘The Haystack’ (see previous post). One possibility is that this was the original route Laurelton Lumber Co. took down to Elk Creek and the Walker Tract, but that it was later abandoned (as work progressed west) in favor of the ledge described above, which might be a bit less steep. Regardless, this ledge also continued up steeply, curved to the south, entered a fair cut through a shale bank, and was rudely interrupted by a sort of road/clearing on the edge of a clear cut area. On the other side of the clearing it appeared the route probably continued, but the thick layer of spice bush made proceeding unattractive. Nevertheless, the direction seemed correct to cross the deer exclusion zone and hook up with previously surveyed grade descending from ‘The Haystack’.
Looking back at the GPS tracks now, I like my theory of the upper ledge being an early route which was replaced by a gentler slope. If you figure both routes climb the same elevation, my proposed later route takes 2-3 times as much length to do it, hence the grade would be substantially less. While the original route might have gotten them started, as they expanded cutting to the west in the Walker Tract it might have been too much of an operational hazard and too difficult to climb with loads.
Note that the areas involved here were part of the ‘Walker Tract’ which Ben Kline’s book describe as “southeast of Livonia”. These were supposedly the last areas cut by Laurelton Lumber and were purchased “in the late 1890s”. The Laurelton sawmill was probably almost 20 miles from here by rail, and crews were probably quite glad when the Walker Tract was exhausted.
According to Rails to Penn State, the Oreland branch of the Bellefonte Central RR was built to haul iron ore from Oreland bank on ‘Whipple farm’ to Bellefonte’s iron furnaces. The branch was removed about 1892, when its rails and ties were lifted to build a 1 mile branch into the town of State College. The college was finally growing into something substantial, and the railroad foresaw more potential for traffic in the borough than it did in a modest ore bank.
Here is an approximate map I of the branch I created by overlaying Penn Pilot aerial photos from 1938 with modern map data. It’s a pain! But I think I have the route down pretty well. The location of Oreland bank is shown; it’s now Lion’s Gate apartments, more or less. I have also marked the location of a stone quarry mentioned in the book, which was apparently where Hamilton Square shopping center is now. In the image below, it appears as a depression at the top right.
If you know the area, it’s a bit bizarre to visualize an early 4-6-0 steam locomotive moving ore and stone through what’s now shopping centers, apartment complexes, and a high school complex.
If you’d like to experiment with an interactive map of the branch, download this gpx file and upload it to GPSVisualizer.com. You can then view several aerial photo formats, as well as google maps, topos, etc. You can also combine this track with others for the Pine Grove Mills branch, etc.
The domain LostRailroads.com has now been registered as a direct way to get to this site. Hopefully it will be easier for you to remember, and more accurately reflect some of the investigations we do, which often extend beyond the logging railroads of central Pennsylvania and the State College region. Thanks for visiting!
For some years I have wondered about the short-lived Pine Grove Mills branch of the Bellefonte Central Railroad, mainly because I used to drive across it several times a day. Some parts of its route are evident, some were less clear. Where exactly did it run? And where was the station?
One time I encountered Cecil Irvin (who operates a farm on Nixon Road and has lived his entire life there) and asked if he remembered the route. He pointed out the fence row along his property. And someone (Mike Bezilla?) pointed me to a suspicious foundation adjacent to the PGM school which might have been the station…
I’m currently reading Rails to Penn State by Mike Bezilla and Jack Rudnicki, and I’m back exploring the branch. Unfortunately, recent renovations at PGM school have paved over the site which Bezilla confirms was the station, and I’m fairly certain the foundation was bulldozed out before the paving.
I’ve created a map based on current and historic aerial views…
To play with a “live” map of the branch, download this .gpx file (it is a just a text file, nothing evil) and upload it at GPSVisualizer.com. You will get a live map which you can pan, zoom, switch between various views, e.g. Google hybrid, USGS topo, etc. If you like GPSVisualizer, please send them a couple bucks!
I’ve shown the approximate locations of the related stations and the Struble ore bank. You may also wish to upload .gpx tracks for the Oreland and Gatesburg branches (see other postings), thus creating an interactive map of the whole region.
On a cold, blustery Saturday with snowflakes trickling through the hemlocks, we decided to return to the Laurelton Lumber operations around Winkleblech and Buffalo mountains at the east end of Centre county.
We started by taking Rt 45 to Sheesley Run Rd., east of Woodward. I wanted to refresh my memory of Sheesley Run Rd, as some aerial views showed possible branches into hollows to the east of the road. A quick review from the truck showed that to be unlikely.
The next point of interest was Buffalo Creek. A study of Kline’s sketch map compared to modern map data suggest that a route must have extended along Buffalo Creek down into “The Gooseneck” and Buffalo Gap. USGS aerial views show something that looks like RoW in the lower reaches of Buffalo Gap, approaching Aichey Rd.
We therefore proceeded up Sheesley Run Rd, across Stony Run Rd onto Buffalo Flat Rd, and over Buffalo Mountain. Stopping at Buffalo Creek, we parked near the adjacent hunting camp and picked up trail to the east. Quickly, Vince declared we were on railroad grade, although I wasn’t entirely convinced. Before long we ran into a thicket where the trail diverged to the right. I pushed ahead through the brush (grumbling!), but soon became convinced that Vince was correct. Under the dense evergreens, snow was lying in clearly defined parallel depressions as the track began to descend towards the stream on a distinct ledge.
Soon, the RoW reaches the stream and an interesting area which well may have been a camp. A lot of rocks were moved around here to some purpose. The small dam on the stream is unlikely to have survived 100+ years, but may mirror an early creation here. At this point, we verified the RoW crosses the stream, and then discontinued our survey. It appears the route would continue at least 3 miles along Buffalo Creek, so a shuttle at the Aichey Rd end seems desirable.
Returning to Buffalo Flat Rd, we looked on the other side of the road, but it appears likely that the road is the RoW over Buffalo mountain. Its route agrees pretty well with Kline’s sketch map, and we can’t detect anyplace along it where right of way diverges. On the map below, we have shown the road as the route until the point where the trail diverges.
Re-crossing Buffalo mountain, we took Stony Run Rd to the northwest to Pine Creek. There, we had previously surveyed a small fragment of RoW, which we theorized was either part of Laurelton Lumber’s operation or Bishop Lumber’s earlier tramroad. Its origin seems unclear: the alignment seems to suggest a connection with Laurelton Lumber, who we know was along Stony Run farther to the south; the construction seems to suggest a tramroad such as Bishop Lumber would have constructed, but in that case its alignment seems wrong…
We decided the most practical approach was to simply reverse our earlier survey and see where we ended up. Parking at the camp just southeast of Pine Creek, we easily picked up the RoW at the end of their driveway. We proceeded to follow it to the northeast. We rather expected we might encounter a junction with Bishop’s “main line” more parallel to Pine Creek Rd, but we never detected anything solid. Rather we crashed through intermittent heavy Mountain Laurel until we had gone approximately a mile, traversing some nice road roadbed with good tie depressions as well as some swampy areas where the route became fairly vague. We then returned to the truck via the adjacent road.
Vince crashed about in the woods near the junction of Stony Run and Pine Creek Roads, and then nominated one final investigation: a drive up the un-gated road to the southwest along Pine Creek, which he thought was a likely route for Bishop’s tramroad. We proceeded about 2/3 of a mile until reaching a bridge that appears structurally deficient. We then proceeded on foot to approximately the end of the road as shown on the map. At that point, the track sort of continues, but doesn’t look much like RoW, even for a tramroad. Scouting around, we detected a more probable route departing slightly to the north of the obvious track. It rapidly enters a swampy area near the stream and is fairly indistinct, but the underlying amount of rock and width seems consistent with a tramroad.
So we still haven’t resolved exactly what’s going on in this area, but it seems most likely that Bishop’s tramroad route hasn’t survived the ages very well, and that Laurelton Lumber later built down Stony Run and over top of the Bishop RoW to head northeast up Pine Creek. How far they might have continued might be discovered by completing a survey along Pine Creek; it’s possible that at some point the RoW deteriorates from “railroad grade” to “tramroad grade”, lending credence to our theory.
We concluded the day with a visit to Elk Creek Cafe in Millheim, where they offer Winkleblink [sic?] Ale, suitable for thirsty logging enthusiasts.
A survey of historic logging (and other) railroads in central Pennsylvania and beyond.