“This can’t be right.” He tapped one of the pages, next to the start of the blue line. “This strain occurred in Buenos Aires only six years after the Rising. It was one of the first signs we had that Kellis-Amberlee was mutating outside a lab setting.”

Those are strain designations, said George. Her voice was very small. Those are the strain designations for some of the most widespread varieties of Kellis-Amberlee.

Everyone has Kellis-Amberlee, but most of us have only one strain at a time. Some are more aggressive than others and will basically wipe out an existing infection in order to take over a body. The original Kellis-Amberlee strain developed when lab-clean Kellis flu met lab-clean Marburg Amberlee. That was the first infection anybody had to deal with, the one that swept the world during the Rising. It took years of study and analysis of the structure of the virus before anyone realized that it was doing what viruses have done since the beginning of time: It was mutating, changing to suit its environment. For a while, people hoped it was becoming less virulent and that it would eventually turn into something that didn’t do quite as much damage. Honestly, I think we’d have been happy if the virus just started killing people, rather than doing what it does now. At least then the dead would stay dead and the world could start moving on. Instead, Kellis-Amberlee has continued doing what it does best: making zombies and unleashing them on the world whenever it gets the opportunity.

I guess it’s consistent. That’s something, anyway.

“It’s correct,” said Mahir. His voice was dark, and there was something dangerous in his tone, something I’d never heard there before. He adjusted his glasses and continued: “There was a spike in deaths in Buenos Aires right before the substrain was isolated and identified for the first time. Eighty percent of the dead were confirmed as suffering from an early form of reservoir condition. It was five years before that substrain was identified in connection with a live reservoir condition.”

Kelly paled further.

“As part of his research into the behavior of the various substrains, Professor Brannon had access to census and death records from multiple parts of the world,” said Mahir. “Much of this data hadn’t previously been incorporated into the model—Dr. bbey is unable to acquire information through many normal channels, due to her lab’s lack of accreditation, Dr. Christopher’s focus is on treatment, not the structure of the virus itself, and Dr. Tiwari doesn’t do statistics.”

“I’m not following you,” I said.

“I am,” said Kelly. She directed her words at the wall, looking faintly stunned. “He’s saying that once they were able to feed the substrain analysis and the census data into the same model, they started getting some results they didn’t want to get. The kind of results a man who spent his life working to save lives would commit suicide over.”

Maggie frowned. “I thought results were sort of the goal.”

“They are, in the general sense, but there are negative and positive results from any analysis. Look at this.” Mahir tapped the paper, shoving it toward Maggie. “Every time a new viral substrain is identified—every time—it comes immediately after a spike in the local death rate. Buenos Aires. San Diego. Manchester. It isn’t a coincidence, and it isn’t confined to any specific country or part of the world. It’s everywhere, and it’s every time.”

Becks shook her head. “What does that prove? Maybe the new strains are more virulent when they’re first getting started, and they’re killing all these people.”

“Unlikely.” He produced another sheet of paper, this one with a brightly colored pie chart on it.

“Eye-catching,” I said, tugging it closer to my side of the table.

“That was the intent.” Mahir pulled another copy of the chart from his file and handed it to Alaric. “This shows the aggregate causes of death among the people with reservoir conditions killed immediately prior to the identification of a new substrain.”

“These wedges are too small to read,” said Alaric.

“My point exactly. There is no dominant cause of death among the victims in these regions. They just… die. They get hit by cars, they fall from ladders, they take their own lives, they die. As if it were any other day, as if theirs were any other deaths. The pattern is in the absolute lack of a pattern, and it’s everywhere, and a month later, there’s a new strain of Kellis-Amberlee running about, more virulent than the one that was in that region prior to the deaths. Three to five years after that, the first reservoir conditions linked to the new strain start showing up, and then it’s another two years before the cycle starts over again.” Mahir removed his wire-rimmed glasses, wiping them on his shirt. “Dr. Connolly, would you care to tell me what conclusions you draw from this data?”

“I can’t make any firm determinations without studying the material more thoroughly, but…” Kelly wiped her eyes with the back of her hand, voice hitching a little as she continued: “I would say there are no naturally occurring viral substrains of the viral chimera generally referred to as Kellis-Amberlee.”

“What are you talking about?” I demanded. “He just said there were new strains appearing all the damn time. This dead professor dude made his career studying them. They ve to exist.”

She didn’t say they don’t exist, Shaun. She said they don’t occur naturally.

Georgia sounded subdued, even resigned, like this was the answer she’d been expecting all along, like the part of me that kept her with me understood perfectly and was just waiting for the rest of me to catch up. I went very still, the skin tightening into goose bumps along my arms as I looked, helplessly, at Mahir. He looked back, waiting. They were all waiting, and they all knew I’d get there if they just gave me a minute. They knew George had the answers, and I… well, I had her.