Question: How do I sustain audience attention for an especially long chapter?
Nothing's wrong with a long chapter. My longest topped out at nearly eight thousand words and some authors write much longer ones. Here's the key: When the chapter is over, the reader needs to know significant headway was made. This does not mean you have to cover a lot of time. A long chapter can cover a few minutes or a few hours, but something has to happen. Nothing's worse than a reader deciding the twenty minutes he devoted to your six thousand word chapter (see Reading Rates) was a waste. I direct your attention to "The Council of Elrond" a chapter of The Lord of the Rings infamous for the amount of time given to what is essentially a group of people sitting around telling stories. Perhaps it's boring, but it works because of one little sentence five paragraphs from the end "'I will take the Ring,' he said, 'though I do not know the way.'" The end of this interminable chapter sets the course for the rest of the book—all 740 pages of it. Even if the reader thinks the rest of your chapter was boring, he cannot deny its impact should something meaningful happen.
(A word about your spectacular ending: Keep the denouement to a minimum. The five paragraphs Tolkien includes after Frodo's inspiring "I will take the Ring," robs his ending of dramatic punch by concluding with these wholly uninspiring words from Sam: "A nice pickle we have landed ourselves in, Mr. Frodo." Tell me, friend, which ending do you prefer?)
But what if your reader never gets to the end? He may feel an obligation to finish (to himself if no one else) if he plunked down $8-$25 to purchase your book. Borrowing it brings with it a time commitment. If he doesn't finish it in two weeks he'll have to return it to the library (assuming it's waitlisted and can't be renewed) or his friend (assuming he's a good friend who returns borrowed items promptly). But your writing needs to be good enough that a reader will finish your book even if he received it for free. I still have a backlog of stripped books I received for free from my days working at a bookstore. Many of them remain on my shelf unread because they failed to capture my interest. Similarly, your book may remain on your reader's shelf—or worse, unsold on a bookstore shelf—and your spectacular ending unread unless you can hook your reader and keep him hooked while you reel him in. How can you do this? I'm glad you asked.
Here are a few hints:
1) Tighten your pacing. Cover a lot of ground. Again, this does not mean you have to cover a long time period. The ground can be emotional. Screenwriters know the best scenes are those that accomplish two or more things at once. A scene in which your protagonist buys a hot dog has to do more than fill his belly. In other words, unless the scene furthers the plot or deepens characterization, cut it. I'd say cut it if it doesn't do both. Your goal as a writer is to tell a complete story, to move your reader from beginning to end with as few detours as possible. Preferably zero. Note that character development is not a detour. To paraphrase bestselling but utterly boring author Michael A. Stackpole, there are plot-based novels and character-based novels. Only the latter sell (to publishers anyway).
2) Preview something big or unexpected. If the beginning of your story is tame, inject a little suspense by previewing a big scene. The big-shot literature profs call this in medias res, meaning "in the middle of things," and it's a technique as old as literature itself. You start the story with your protagonists in deep trouble. Then you flash back and show how they got there. There's an old stomach-churning phrase used to describe that kind of trouble: "When the shit hits the fan." When you start your story in medias res, you start with the shit an inch or so from the fan. You flash back to when it was a foot away and spend the rest of the story following the shit's trajectory back to one inch and beyond to where it splatters and spreads (I told you it was stomach-churning).
In some ways I consider Mizuho Mishap (The Swordsman and the Summoner chapters 12-17) one long chapter. I knew I wouldn't get to any significant action until later in the story (and no real action had occurred in the first book either). So I began with a scene that presaged a knock-down-drag-out between two former allies. Unexpected? You betcha. And for the next several pages the reader anticipates an explanation for how the two came to be at each others' throats. And he will keep reading until he has it. The resolution comes in time (towards the end of chapter 14) but by then I have presaged other fights, other conflicts to keep the reader interested. And that's hint three...
3) Never answer one question without raising another. Or as Stackpoles put it, give your reader what he needs, not what he wants. He wants full resolution and he wants it NOW, damn it! But as soon as he gets what he wants, he will stop reading. If Hero and Heroine kiss during the first ten pages, that kiss is no milestone. If it happens ten pages from the end, it very well may be. (On a related note, what was Pete Jackson thinking having Adrien Brody and Naomi Watts kiss before they even reached the island in King Kong? He rendered the whole romantic subplot moot before the audience met the titular character. This storytelling gaffe irritated me so much I couldn't finish the film. Seriously, I didn't make it to Kong's first scene.) Once your central conflict (hereafter referred to as CC) is resolved, your story is over. You may go on writing for another hundred pages, but your story is over. This CC is not necessarily the one you introduce on page one. Feel free to introduce a few false ones along the way, but don't build them up so much that your reader feels he's reading a different story after the false CCs are resolved. Try relating the false ones to the real one. Frodo thinks his purpose is to get the ring to Rivendell. So does the audience. Both are wrong. But transporting the ring to its final destination is the CC. Swordsman's CC seemed to be the Iselia invasion. This conflict was built up so much that one reviewer erroneously believed the story was nearing its end in chapter fourteen (Technically that's probably my fault. The action was escalating at a frantic pace indicative of a story climax. But in my defense I did introduce a new conflict before the Iselia defenses were breached). This brings up a final point: Your true CC is the one recognized as such by the majority of your audience. Your intended CC is beside the point if it does not match your audience's interpretation. If I had stopped writing at the end of book two, the CC would be the invasion whether I meant it to be or not.