Two days later I was back at work, re-designing the first drop as a horseshoe that gently swooped down into a double-dip at the second hill. Though I re-used as many of the old bent sections as I could, I still had to construct seven new ones. I decided to get rid of the two rear ticket booths and move the Ferris wheel to the back of the float bed. I had about a month until the next parade, the day after Thanksgiving. I wanted the float operational by then.

To power the lift, I designed a simple pulley system. A spring-loaded arm underneath the float deck would have a wheel attached to it. The spring would keep the arm pulled down toward the ground. Attached to the wheel would be a pulley. A belt would go up through the float deck, to another pulley on a shaft near the Ferris wheel. That shaft would have two other pulleys on it: one to drive the Ferris wheel and the other under the lift hill. A belt would extend to another pulley on a shaft at the start of the lift hill. Attached to this shaft would be the "chain." Instead of a chain, I would use fabric belt similar to those used on flume rides. It would rub against the bottom of the train, pulling all the cars over the top of the lift. There was one catch to these grandiose plans: because of other duties I had at the college, I really would have only two weeks to rebuild the float. Carcus had other commitments and was unavailable to help me. I was on my own.

I hurridly assembled and painted the new bent sections, and that's about all I accomplished in a week and a half. Three days before the Thanksgiving parade, I returned to the garage and inspected the float. Because of the soaking it got in the previous parade, there was some minor buckling of the wood but surprisingly no major damage. I trashed the two rear ticket booths. I dismantled the coaster structure from bents ten through eighteen. In one day, I rebuilt the entire first drop and re-tracked it. I couldn't re-use the old track; the new bent heights made that impossible; the new structure didn't drop as far down to the deck as the old one, and there was less track length needed. Instead, I overlapped short pieces of new track, from ledger to ledger. I also pulled the track in about a half-inch from each side to prevent the train from falling through. Then I detached the Ferris wheel and moved it to the back center of the deck.

The next day, I added new side-friction rails. This time, they were made from 1/4" plywood for durability. I also pulled in the lift hill track by a half-inch on each side. The final section of track to be attached was between the top of the lift and the beginning of the first drop. Once again, it was too abrupt a transition and I couldn't bend the plywood to fit it. I should have made that bent section about five inches shorter. So I created an awkward butt joint and sanded it down so that it was smooth. The result was certainly not ideal. The bottom of the train scraped against the ledger as it tried to negotiate the v-shaped hump. The parade was going to be televised, and I wanted the train running. But I realized at that point that there was no time to put together the lift mechanism. So I decided that I would simply pull the train to the top of the lift by hand and just drag it over the spike in the track.

Finally the revised coaster was completed. I hesitantly pulled the train to the top of the lift, held my breath and let gravity take over. Slowly at first, the train rolled around the first curve, gained speed and then barreled into the double-dip. It sped down the second drop and nose-dived into the track, ripping up a chunk of it and stopping dead. After mumbling to myself a bit, I tore up that section of track (where I had used luan strips). I rebuilt it with strips of 1/4" plywood and lessened the angle of descent so that the train couldn't "bottom out."

I pulled the train once again to the top of the lift and released it. The train thundered forward, hopped over the double dip and dove down the second drop. The train roared up the turnaround and picked up a little air time, enough to make it hop over the outer side-friction rail, collide with the top of bent number twenty-six and snap it in half like a brittle twig.

I pulled out of the dumpster two of the bent sections I'd thrown away. Within an hour I had modified them to replace bents 25 and 26. I was very glad that I put the structure together with multi-purpose screws; that made quick modifications like this possible. In short order, I was ready to attempt another ride.

I held my breath tensely and released the train. It rolled down through the double-dip and gracefully negotiated the turnaround, rumbling back to the lift and stopping. I let out a whoop. I had finally succeeded.

Rain fell from the sky for Springfield's "Parade of Big Balloons." I wasn't about to run the coaster in a downpour, so I propped the train at the top of the lift and let it sit idly there. I rode in the truck cab with Tom Stewart. As we neared the reviewing stand, the rain stopped. I got out of the cab and hopped up onto the float. I gave the train a little push, and it rolled down perfectly through its course and made it back to the lift. The crowd cheered. I pulled the train back up the lift and sent it on its way again. This time, as it reached the bottom of the second drop, the train did a nose dive and ground to a halt. That crash tore up the front metal trim on the first car. I patched it up and pushed it over to the lift. I was determined to have the train rolling for the television cameras at the reviewing stand. On subsequent runs, the train stalled out right before the turnaround. But I just pulled it back up the lift and ran it again. The spectators were thrilled, and the commentators commended the creativity of the college.

The problem was in the couplings. Each time the train ran its circuit, the malleable metal perf would bend and twist causing each car to point its front down. Each time I brought the train to the top of the lift, I had to straighten out every coupling.

The train took a nose-dive at the bottom of the second hill again and tore up the inner side friction rail there. Luckily, the float was past the reviewing stand by then. I pulled the train to the top of the lift and tied it off. I spent the rest of the parade route by the Ferris wheel, turning it slowly by hand and waving to the crowd.

The next day I bought 5/16"-thick eyebolts to replace the metal perf couplings. I attached them to the 2x4 underneath each of the cars. Simple nuts and bolts with large washers served as pins to hook the couplings together. By this point, the track had begun to warp more severely from the rain. The thin side-friction rails especially were bowed in and out. The Ferris wheel looked like an unfinished mess (because it was). The train wheels had begun to show signs of wear, just from their small amount of use. The next parade was the big Holyoke St. Patrick celebration. It was about four months away.

The winter quickly passed and all I accomplished was the reworking of the train couplings and the reconstruction of some of the side-friction rails. Carcus reappeared and I had him put together graduation caps out of cardboard, hot-gluing the pieces together. I used pieces of fringe for the tassles. Then I had him paint the train gold and green, the school colors.

Two weeks before the parade, I stood on the deck of the float and pulled the train to the top of the lift. I let it go, and it crawled forward slowly. It lumbered past the double-dip, down the second hill and collided headfirst with the ledger at the bottom of the hill. The transition there was still a bit too abrupt. The thick eyebolts didn't flex as much as the metal perf used to. The runners underneath the first car (to which the side-friction wheels were attached) were catching on that one ledger. I took a rasp and whittled away at the ledger to lower its profile. Then I took the rasp to the runners themselves. I pulled the train to the top of the lift again and let it go. This time, the train made it past the ledger but once again slammed into the side-friction rail before the turnaround and broke it. I rebuilt that rail with 1/4" plywood and tried again. When the train got to that turn, it had picked up enough speed to become airborn. The front car flew off the track and hung off the edge like a dog with its head out a car window.

I picked up the train and discovered that the eyebolts were bent. I had underestimated the incredible amount of force the train had. I straightened the bolts, took some leftover metal perf and made collars for the eyebolts so that they couldn't bend.

I pulled the train up the lift and let it go. Once again, the train got to the final turnaround and flew off the track. The side friction rail there needed to be higher. So I raised it. I ran the train again, and this time the front car's side-friction wheel got jammed underneath the side-friction rail. So I lowered the rail a bit and tried again. The train made it to the top of the turnaround...and stalled out. I ran the train over and over, and it kept stalling out at that same point. I watched the train carefully as it ran, and I noticed that the last car was bouncing all over the track like a ping-pong ball. So I detached car four and made the train shorter. I ran the train again, and this time it barely crested the turnaround and very slowly rolled back to the lift. The last car was still bouncing around, but much less so. I tried flipping the couplings 90 degrees, to give the cars more vertical movement. The result was still the same.

Then it dawned on me: the eyebolts were connected to the 2x4 of each car, which threw the car's center of gravity off. The cars were designed to be towed from the front, and I was trying to pull them from underneath their center. I needed a new type of coupling--fast! I wracked my feeble, tired brain, trying to think of some way to quickly and cheaply engineer a small u-joint (which would allow for both lateral and vertical movement). A simple idea occurred to me. I rushed to the hardware store and came back with more eyebolts. But this time I bent them at ninety-degree angles, with the hole parallel with the bend. I attached the bolts through the front and rear of each car and double-nutted them in place. Then I re-used the bolts I had for the pins. Voila! I had a simple u-joint. I dragged the train to the top of the lift and let it go. It seemed to articulate much better. But as it reached the bottom of the double dip, it slowed to a crawl and stopped.

Carcus noticed that each car was "hunting" a little more. The metal trim on the front of each was scraping against the side-friction rails. In some places, the nose of each car was making contact before the side-friction wheels did. So I ripped off all of the metal trim. Then I had Carcus cut more of a bevel into the front of each car, which exposed more of the side-friction wheels. I applied new trim made out of pieces of the ribbon bracing. I figured that would cause less friction than the metal trim, even though it didn't look as nice. Actually, I should have forgone the trim entirely. That would have eliminated the scraping problem.

I once again pulled the train up the lift. It thundered smoothly through the entire circuit. In fact, for five consecutive times the train made a perfect circuit. I was confident it would run fine for the parade. But I knew with the time remaining that I would again fail to construct a working lift hill mechanism.

I set Carcus to work painting the Ferris wheel: yellow for the wheel and green for the supports. We placed the "graduation caps" on it. I drilled holes in the hub shaft and fitted cotter pins into it to keep the wheel from drifting left or right as it turned. The paint job spruced it up quite a bit. Carcus added some garland to finish it off and the Ferris wheel was done.

A couple days before the parade, we added some more garland to the remaining ticket booth and swept the float deck. I stood back and looked. This was it.

Finally, we had a beautiful day. The Holyoke St. Patrick's Day parade took place under sunny skies and mild temperatures. There was a four-hour wait at the parade site before things got under way, so I took a lot of video footage of the coaster and tested the ride over and over. The cars had enough movement to them now that occasionally some would turn a little too much and a wheel would drop through between the tracks. Other than that, it ran flawlessly until a few minutes before the parade began: the coupling on train four sheared in half. It was yet another reminder of the incredible forces at work on a coaster. I hurridly detached that car, and was left with a shorter train (which fortunately still ran fine). As we progressed along the parade route, the densely-packed spectators often called out to me, asking if it was the old Mountain Park coaster.

We fit five people on the float in addition to myself. I assigned one, a thirteen-year-old boy, to operate the Ferris wheel. He did a great job throughout the long parade, making it appear as if the wheel turned all by itself. I walked back and forth, repeatedly bringing the train to the top of the lift and letting it go, then meeting it at the base of the lift and starting all over again. It sporadically got hung up on the turnaround, but I just pulled it on through and back to the lift.

I noticed an unexpected phenomenon: there was a steep hill we had to negotiate along the route. As we went up the hill, the train kept stalling out in the middle of the double-dip. The Ferris wheel boy would give it a push down, and then it would stall out again at the turnaround. The hill had the same effect as tilting the whole structure back. Then as we went downhill, the train flew through the course so fast it came off the track at the turnaround.

At the previous two parades, the announcers at the reviewing stand pointed out the graduation caps (which at that time didn't exist) on the Ferris wheel. Ironically, now that we had the caps no mention was made of them. But it was nice to hear the warm reaction from the crowd. They knew what it was and they seemed to like it.

When I stood back from the float, it looked so tiny. It was hard to believe that such a relatively small structure took so long to build. All in all, I was pleased with how the project turned out. As a bonus, it won third prize in the parade's Open division.  The float wasn't destroyed immediately after the parade. When the time came for the college to strip the deck down and begin work on their next creation, I dismantled it for them. With the help of my sister's boyfriend, I brought the pieces home and reassembled the coaster in our backyard where it surrounded a little garden that Karen cultivated.

I wasn't happy with some aspects of the float (not making a lift mechanism, for instance). But I had done what I set out to do: create a working full-circuit coaster. Along the way I made a lot of mistakes, and from them taught myself some important lessons in coaster-building. Now I'm ready to tackle a "real" one! Learning is indeed life's biggest thrill.