“Yeah, because that would work.”

“It might work to get you in,” I said. “He’d promise you an interview for sure.”

She gave a snort. “That’s not actually a stupid idea. Except I don’t want to talk you into New York, I’m just…” she trailed off. “El, it’s not just Magnus,” she said bluntly. “I’ve had a bunch of people asking me. Everyone knows about you now. And if you don’t go to some enclave—they’re going to wonder what you are planning to do.”

And I didn’t even need to lay my idiotic plans out before her to know without question that they’d believe in them even less than she would.

* * *

To further gladden my heart, our midterm grades were coming in that week. No matter how hard you’ve worked, there’s always something to worry about. If you’re going for valedictorian, anything less than perfect marks is a sentence of doom, and if you do get perfect marks, then you have to worry whether your courseload is heavy enough that your perfect marks will scale well against all the other kids going for valedictorian with their perfect marks. If you’re not going for valedictorian, then you have to spend the bare minimum of time on your actual coursework in order to maximize whatever you are actually working on to get you through graduation—whether that’s expanding your spell collection or creating tools or brewing potions, and of course building mana. If you get good marks, you wasted valuable time you should have spent on other things. But if your marks are too bad, you’ll get hit with remedial work or worse.

If you’re wondering how our marks get assigned when there are no teachers to evaluate anything, I’ve heard a million explanations. Loads of people, mostly enclavers, say with great assurance that the work gets shunted out of the Scholomance and sent to independent wizards hired for the marking. I don’t believe that for a second, because it would be expensive, and I’ve never met anyone who knows one of these wizards. Others claim the work gets graded through some sort of complicated equation based almost entirely on the amount of time you spend on it and your previous marks. If you want to really set off any valedictorian candidate, try telling them that it’s partly randomized.

Personally, I’m inclined to think we’re doing the marking ourselves, just because that’s so efficient. After all, we mostly know what marks we deserve, and we certainly know the marks we want to get and what marks we’re afraid of getting, and whenever we see bits of someone else’s work we get an idea of what they ought to get. I’d bet that the school more or less goes by the sum of the parts, depending on how much will and mana each person has put behind their judgment. Which also handily explains the bloc of self-satisfied enclavers who annually fill up the class rankings only a little way down from the actual valedictorian candidates, despite not doing nearly as much work and not being nearly as clever as they think they are.

None of these possibilities told me what to expect in the way of marks for myself, since this year I was entirely alone in the one seminar to which I’d devoted massive amounts of time and energy and passion and mana, and apart from that I was in three other small-group seminars that I’d aggressively neglected.

You’d think that marks wouldn’t matter very much senior year unless you were going for valedictorian, since we literally don’t take classes for the second half of the year. After the class standings are announced at the end of the first semester, seniors are given the rest of the year off to prep for their graduation runs.

But that’s in the nature of a grudging surrender. Graduation wasn’t designed to be a slaughterhouse gauntlet of mals. The cleansing routine we fixed last term was intended to winnow them down to a reasonable level every year before the seniors got dumped in to make their escape back to the real world. After the machinery broke four times in the first decade, and the enclaves gave up on repairing it, seniors largely stopped going to class, because there’s a point where training and practicing with the spells and equipment you’ve got is more important than getting new ones. When there’re a thousand howling starved mals coming at you from all sides, you want your reactions worked thoroughly into your muscle memory.

So the powers that be running the school at the time—London had taken it over from Manchester by then, with substantial support from Edinburgh, Paris, and Munich; opinions from St. Petersburg, Vienna, and Lisbon taken under advisement; New York and Kyoto occasionally given a patronizing ear—decided that they’d accept the reality and turned it into a deadline. And up to that deadline, the school does its best to make marks matter. The penalties get especially vicious for our last semester. Finals are the worst of it, but even midterms are generally good for taking out at least a dozen seniors.

I was relatively safe for the Proto-Indo-European seminar because I’d cheated on that one essay, which always gets you good marks. The school’s perfectly willing to let you leave dangerous gaps in your education. The only risk there was that I’d actually done some work on the thing beyond copying it out; I’d get marked down for that, although probably still not to failing level.

The translation I’d handed in for my final Myrddin seminar poetry assignment was a rotten half-baked job I’d run through in two hours, guessing wildly at the many words I didn’t know. It hadn’t left me with any spells I could even sound out, at least not if I didn’t want to risk blowing the top of my own head off. I’d got top marks on the lovely deconstruction spell I’d used on Chloe’s cushion-monster, though, which was likely to pull my grade there to safe levels.

Aadhya had helped me with the maths in my Algebra course, and I’d done a load of translations for her. Artificers don’t get language classes in their senior year; instead they just get assigned a design project in one of their other languages. Design projects are a really special fun thing for artificers. You’re given a set of requirements for some object, you write down the steps to build the thing, and then the Scholomance builds it for you. Exactly according to your instructions. Then you have to try the resulting object out and see what it does. Three guesses what happens if your instructions turn out to have been wrong or insufficiently detailed.

Having to do a design project in a second language makes it even more exciting. And in this case, Aadhya’s languages are Bengali and Hindi, both of which she knows really well, except the school swerved and gave her the projects in Urdu instead, which happens sometimes if it’s feeling particularly nasty. She didn’t know the script well, and anyway subtle differences in meaning matter quite a lot in these circumstances. You’d really like to be confident that you’re not building a blasting gun backwards, for instance.

You would also really like to build something that might be of some use at graduation, but her choices were a mana siphon, a shell-piercer, and a garden planter. A mana siphon is a flat-out maleficer tool and anyway the last thing anyone allied with me would ever need. A shell-piercer would’ve been a terrific weapon against constructs, except in the very last line of the specifications, the assignment said that the purpose of this particular one was to acquire usable miercel shells. Miercels are these self-reproducing construct mals that look rather like wasps the size of my thumb. Their shells are made of a mana-infused metal and are quite useful, but a shell-piercer of any combat-appropriate size would fracture them to tiny bits.