ANIMAL CRACKERS all broken up at the bottom of the box … "cracked," you might even say! Clever interpretation of CRACKERS, meaning that some black squares crack, or divide, an ANIMAL in two (see below, highlighted).
![Typical ANIMAL CRACKERS. Really, a sheep? Huh. Typical ANIMAL CRACKERS. Really, a sheep? Huh.](https://upload.wikimedia.org/wikipedia/commons/thumb/f/f2/Barnum%27s_animals_examples.JPG/600px-Barnum%27s_animals_examples.JPG)
Even better was that Herre stuck to long-ish animals typically seen in boxes of ANIMAL CRACKERS! Would have been easy to pick short animals like HEN or FROG or ELK, but that would have felt odd. Speaking of odd, not sure why it's unappetizing to eat a hen-shaped cracker vs. a ZEBRA or an ELEPHANT. I call fowl! Er, foul.
Well, it's not that difficult to split up semi-long words so that half ends an entry, and the other half starts another. Even a long guy like GORILLA gives multiple possibilities for *GOR and ILLA*. The latter not so much, but still, with I'll ask, ill-advised, ill at ease, etc., that's a good amount of flexibility.
BUT! Even with that flexibility, it's no joke to work around five pairs of themers. PHANT* doesn't give many possibilities, and when you must obey crossword symmetry, things can get hairy. I thought Herre did extremely well in the top half of the grid, working in some beautiful OH COME ON, NO GO AREA, I HOPE NOT bonus fill — all without crossword glue! Some might complain about LON Chaney, but he was a famous actor. Fine by me.
The bottom half didn't come out as nice, what with AMO, RCMP, AJOKE, LSTS, SYS — that's almost too much for one puzzle, and certainly too much for half of a puzzle. It's a much more difficult section to fill since CRACKERS is fixed into place, whereas the top had much more flexibility (MCGREGOR could have been dozens of other things). Along with the fact that Herre had to spend many of his black squares in the middle of the puzzle, sometimes you just have to accept that there will be compromises like this.
Herre could have gone down to just four pairs of themers, but only three animals would have felt light. Although the SE wasn't very elegant, I think it was a reasonable trade-off to get the higher theme density.
Given how well the theme was obscured, my solving experience played out like a pretty smooth themeless, and then the a-ha at the end was delightful — so much so, that I was able to overlook the gluey SE. Such a fun experience overall!