The first attempt failed. The second was worse. The third showed enough improvement to make me believe, and that is how it happens: not with certainty, but with small reasons to go on.
Her temperature came down. Her breathing eased. Her blood markers began to move in the right direction. The platform we had spent years building had never been designed for this exact threat. It worked by pressing on the pathogen from many sides at once, and we had hoped that would be broad enough. Now I needed it to be. I changed the strategy, not the science underneath it but the timing, the combination, the way we watched the response. I worked from memory when the records were not enough. I called colleagues. Some answered, and some did not, and I barely noticed which.
On the second day Anselm Roth flew in from across the country to help. He was the best clinician of infectious disease I knew, and he spent most of an afternoon with me in the isolation corridor. He shook hands with the nurses and drank bad coffee from the same machine as everyone else, and that evening he flew home to prepare his own hospital in case I turned out to be right about how the thing could go. I was grateful to him. I have not stopped thinking about how grateful I was.
For two days I slept in chairs. On the third I fell asleep with my hand on the laboratory table, and Lena woke me.
"You need to go home."
"I don't want to."
"Daniel, you've been awake almost sixty hours."
"I know."
"Then you know you're making mistakes."
"I'll check them."
"You already did." She turned my notes toward me. In the middle of a line of calculations, in my own handwriting, I had written the same number twice. I crossed it out. Lena took the pen out of my hand.
"My mother is dying."
"I know."
"So I can't."
She stood there a moment. Then she set a paper cup beside my elbow. "Tea."
"I don't want tea."
"I know. She would tell me to make you drink it."
I took the cup. It was cold, and I drank it anyway.
On the fourth day her fever broke. I was standing outside the isolation suite when a nurse read the result twice and looked up at me. "That's good," she said, and I could not answer, so she said it again. "That's good."
Through the glass my mother was awake, and she lifted one eyebrow. I opened the communication line.
"Your fever is down."
"Good."
"Way down. The treatment is working."
"Don't get excited."
"I believe I'm allowed."
"No, you aren't." But she was smiling. "How long have you been awake?"
"I don't know."
"Daniel."
"Three days."
She stared at me. "Go to sleep."
"I will."
"Now."
"I can't. I have to watch you."
"I'm not going anywhere," she said, quite casually. I heard it then as reassurance. Later I heard it differently.
The next morning she ate half a bowl of soup and complained that it was terrible. The nurse apologized. "It tastes," my mother said, "like warm water that has heard a rumor about chicken." The nurse laughed, and I laughed, and my mother looked almost healthy, her color back, her breathing easy. She asked for the crossword, and I brought it down to her.
"You didn't solve yesterday's."
"I was busy."
"You've become lazy."
"I'm saving the world."
"You always have excuses." She frowned at the page. "This clue is wrong."
"The clue isn't wrong. You just don't know the answer."
"I know the answer."
"What is it?"
She smiled. "Ask me tomorrow."
"Why?"
"Because I want to annoy you." She set the pencil down. "Daniel. Come closer."
I stepped up to the glass, and she lifted her hand, and I pressed mine against it. She looked at our hands for a while. "You're all right?"
"Yes."
"Really?"
"Yes."
"Good." She closed her eyes.
I stayed there. For the first time since that night in the cold room, I let myself believe she would live.
That was my mistake. Not the hope. The certainty.